School of Computing and Engineering

Pearson Higher National in Computing

Assignment Brief - BTEC (RQF)
Academic Year 22-23

Student Name: Alishba Khalid Class: HND-Computing Y2

Assessor Name: Unaza Tallal Issue Date:
Unit No. and Title: Unit 26: Machine Learning
Assignment 01: Customer Engagement Prediction using Machine
Assignment No. and Title:
Learning Techniques in a Marketing Company
Formative Submission
30th March Submitted on: 30th April
Date:
Summative Submission
3rd May Submitted on: 14th May
Date:
Resubmission Date: 16th May Submitted on: 28th May
IV of Brief Name: Date of IV: 21st May 23

Submission Format
This assignment can be either hand-written or typed in full, but all working must be shown in
order to demonstrate your understanding of the tasks.

This assignment is to be submitted directly to Moodle on the date shown above.

Unit Learning Outcomes
Analyse the theoretical foundation of machine learning to determine how an intelligent
LO1
machine works.
LO2 Investigate the most popular and efficient machine learning algorithms used in industry.

Student Declaration

I certify that the assignment submission is entirely my own work and I fully understand the
consequences of plagiarism. I understand that making a false declaration is a form of malpractice.

Pearson BTEC HN RQF Assignment Brief and Student Declaration Form Template Academic Year 2022/23
 Student Signature: Date:

Attach this section to the front of your completed work

Pearson BTEC HN RQF Assignment Brief and Student Declaration Form Template Academic Year 2022/23
Assignment Brief
Vocational Scenario
You have just been hired as a data analyst for a marketing company that is looking to improve its
customer engagement strategies. Your manager has tasked you with using machine learning
techniques to analyze customer data and identify patterns that can be used to improve customer
engagement. Specifically, you will be analyzing data related to customer demographics, purchasing
behavior, and customer feedback.
Assignment Activity and Guidance
Task 1: Data Preparation Your first task is to clean and prepare the data for analysis. The data is
stored in a CSV file and includes information on customer demographics, purchasing behavior, and
feedback. You will need to perform the following tasks:

1. Load the data into a data frame using a Python library like Pandas.
2. Remove any duplicates or irrelevant data.
3. Check for missing data and decide on the best way to handle it (e.g. imputation, deletion, or
other methods).
4. Transform any categorical data into numerical data using techniques like one-hot encoding
or label encoding.

Task 2: Exploratory Data Analysis Once the data is cleaned and prepared, your next task is to
perform exploratory data analysis to better understand the relationships between the different
variables in the data. You will need to perform the following tasks:

1. Use visualizations like scatterplots, histograms, and heat maps to identify patterns and
correlations in the data.
2. Identify any outliers or anomalies in the data that may need to be handled differently.
3. Use statistical measures like correlation coefficients and hypothesis tests to identify
significant relationships between variables.

Recommended Resources
Please access HN Global for additional resources support and reading for this unit. For further
guidance and support on report writing please refer to the Study Skills Unit on HN Global. Link
to www.highernationals.com

Pearson BTEC HN RQF Assignment Brief and Student Declaration Form Template Academic Year 2022/23
Learning Outcomes and Assessment Criterion
Pass Merit Distinction
LO1 Analyse the theoretical foundation of machine Critically evaluate why machine
learning to determine how an intelligent machine works learning is essential to the
Analyse the types of learning design of intelligent machines.
Evaluate the category of
problems.
machine learning algorithms
Demonstrate the taxonomy of
with appropriate examples.
machine learning algorithms.
Investigate the most popular and efficient machine
learning algorithms used in industry
Investigate a range of machine
learning algorithms and how
these algorithms solve the
learning problems.
Analyse these algorithms using
Demonstrate the efficiency of
an appropriate example to
these algorithms by
determine their power.
implementing them using an
appropriate programming
language or machine learning
tool.

Pearson BTEC HN RQF Assignment Brief and Student Declaration Form Template Academic Year 2022/23
Formative Feedback

Student Name:

Assessor Signature: Date:

Student Signature: Date:

Summative Feedback

Student Name:

Assignment Grade: Choose an item. Assessors Signature:

Resubmission Feedback

Final Assignment Assessors Signature:

Choose an item.
Grade: Student Signature:
Please note that grade decisions are provisional. They are only confirmed once internal and external mode
have been agreed at the Examination Board.
IVs Signature and Date:

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Table of Contents
Assignment Brief...................................................................................................................................... 2

Formative Feedback.................................................................................................................................4

Summative Feedback............................................................................................................................... 4

Introduction............................................................................................................................................. 5

LO1 Analyse the theoretical foundation of machine................................................................................6

learning to determine how an intelligent machine works.......................................................................6

“P1. Analyze the types of learning problems.”........................................................................................6

1. Machine Learning:........................................................................................................................ 6

2. Types of Machine Learning:.......................................................................................................... 7

3. Learning Problem in Machine Learning:........................................................................................7

4. Common machine learning problems faced while analyzing customer data and improving
customer engagement......................................................................................................................... 8

P2. “Demonstrate the taxonomy of machine learning algorithms.”........................................................9

5. Taxonomy of machine learning:....................................................................................................9

P3. Investigate a range of machine learning algorithms and how these algorithms solve the learning
problems................................................................................................................................................ 15

 Logistic Regression:.....................................................................................................................17

P4. Demonstrate the efficiency of these algorithms by implementing them using an appropriate
programming language or machine learning tool..................................................................................21

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
M1. Evaluate the category of machine learning algorithms with appropriate examples......................30

M2. Analyze these algorithms using an appropriate example to determine their power.....................31

D1. Critically evaluate why machine learning is essential to the design of intelligent machines...........36

6. References:................................................................................................................................. 39

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Introduction

As a data analyst at a marketing company, my primary objective is to enhance customer engagement

strategies by harnessing the power of machine learning techniques. Through the analysis of customer
data encompassing demographics, purchasing behaviour, and feedback, my aim is to uncover valuable
patterns and insights that can drive improvements in customer engagement. To achieve this, I will
delve into the realm of machine learning, understanding its diverse challenges and exploring a
taxonomy of algorithms. Python and Google Collaboratory will serve as my trusty tools, enabling me
to explore and evaluate a wide range of machine-learning techniques effectively. By the culmination
of this project, I will have thoroughly assessed machine learning as a class of algorithms, scrutinized
their efficacy through pertinent examples, and offered a critical evaluation of the significance of
machine learning in the advancement of intelligent machines.

LO1 Analyse the theoretical foundation of machine

Learning to determine how an intelligent machine works

“P1. Analyze the types of learning problems.”

1. Machine Learning:

Machine learning is a branch of artificial intelligence (AI) that focuses on the development of
algorithms and models capable of learning from data and adapting to new circumstances. These
algorithms utilize statistical and mathematical models to analyze data, identify patterns, and make
predictions or decisions without explicit programming. The ultimate objective of machine learning is
to create algorithms that enhance their performance over time by learning from previous experiences
or data, eliminating the need for explicit instructions. Machine learning techniques have found
applications in various domains such as computer vision, speech recognition, natural language
processing, and recommendation systems. There are three main types of machine learning:
supervised learning, unsupervised learning, and reinforcement learning. In supervised learning, the
algorithm learns from labeled data, while unsupervised learning focuses on discovering hidden
patterns in unlabeled data. On the other hand, reinforcement learning involves learning through trial
and error, with actions taken to maximize rewards. Machine learning has gained significant popularity

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
and has become indispensable in the business and industrial sectors due to its ability to enhance
decision-making, improve efficiency, and enable the development of new applications and services.

2. Types of Machine Learning:

 Supervised Learning: Supervised learning involves the algorithm learning from labeled data,
where the training dataset consists of input and output variables that provide the correct answers.
The algorithm's task is to establish patterns and relationships within the data, mapping the input
to the expected output. Through this process, the algorithm gains the ability to generalize its
knowledge to new inputs. To evaluate its performance, the trained algorithm is tested against
fresh data. Supervised learning encompasses various techniques such as regression, classification,
and decision trees, which further contribute to the algorithm's ability to make accurate
predictions or classifications based on the provided labeled data.
 Unsupervised Learning: Unsupervised learning involves the algorithm's attempt to discover
hidden patterns within unlabelled data, where there is no known output provided. The algorithm
is presented with a dataset consisting solely of input data and its task is to identify inherent
relationships and patterns within the data. The output generated by the algorithm is typically a set
of structures or clusters that represent the identified patterns within the data. Unsupervised
learning encompasses various techniques such as clustering, dimensionality reduction, and
anomaly detection, which aid in revealing meaningful insights from the unlabelled data without
any prior knowledge of the expected outcomes.
 Reinforcement Learning: Reinforcement learning is a learning approach that operates through a
trial-and-error process, focusing on maximizing a reward signal. Unlike other learning methods,
reinforcement learning does not rely on predefined input/output pairs. Instead, the algorithm
receives feedback in the form of rewards or punishments based on the actions it takes within its
environment. Through repeated trials and interactions, the algorithm learns to make decisions
and take actions that optimize the received rewards. Reinforcement learning finds extensive
applications in areas such as robotics, gaming, and various other domains where agents learn and
improve their performance by actively engaging and interacting with their environments

3. Learning Problem in Machine Learning:

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Each type of learning problem requires a different approach and a different set of algorithms. The
choice of which type of learning problem to use depends on the nature of the data and the task at
hand. Here are some common learning problems in machine learning,

 Classification: This task involves making predictions of a categorical output variable by

considering certain input variables. An example of this is the classification of an email as either
spam or not spam, based on the content of the email. In such cases, the algorithm analyzes the
given input variables and applies a classification model to assign the appropriate category to the
email, thus determining whether it is classified as spam or not.
 Regression: This task involves making predictions of a continuous output variable by considering
specific input variables. For instance, it entails predicting the price of a house based on various
features like its location, size, and the number of rooms it possesses. In this scenario, the
algorithm examines the provided input variables and employs a regression model to estimate the
house's price. By considering factors such as the house's location, size, and number of rooms, the
algorithm aims to generate an accurate prediction of its market value.
 Clustering: This task involves categorizing or grouping similar data points together, primarily
based on their similarity to one another. For instance, it could entail grouping customers based on
their purchasing behaviour. By analyzing various aspects of customer behaviour such as their
buying patterns, preferences, or demographics, the algorithm aims to identify distinct customer
segments or groups that exhibit similar characteristics. This enables businesses to gain insights
into their customer base and tailor their marketing or service strategies accordingly.
 Dimensionality Reduction: This task involves minimizing the number of input variables while
retaining crucial information. For instance, it may entail reducing the number of features in an
image while still preserving its content. Through various techniques and algorithms, the goal is to
eliminate redundant or less significant features from the image representation while maintaining
its essential visual information. By doing so, the algorithm aims to reduce the complexity and
dimensionality of the data, making it more manageable for subsequent analysis or processing,
while still retaining the core content or meaningful aspects of the original image.
 Anomaly Detection: This task involves identifying and highlighting outliers or anomalies within a
dataset that deviate from the expected patterns. For instance, it can be applied to detect
fraudulent transactions within a financial dataset. By analyzing various data attributes and
patterns, the algorithm aims to identify instances that significantly differ from the norm or exhibit

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 suspicious behaviour. This allows for the identification of potential fraud or irregularities within
the dataset, enabling businesses to take appropriate actions for further investigation or
mitigation.
 Recommendation Systems: This task involves providing personalized recommendations to users
by considering their preferences and behaviour. For instance, it could involve suggesting movies or
products to a user based on their viewing or purchase history. By analyzing the user's past
interactions, such as the movies they have watched or the products they have purchased, the
algorithm aims to understand their preferences and make relevant recommendations. This
personalized recommendation system utilizes techniques such as collaborative filtering, content-
based filtering, or a combination of both, to suggest items that are likely to align with the user's
interests. By leveraging historical data and user behaviour, these recommendations enhance the
user experience by offering tailored suggestions that match their preferences.
These are just a few examples of learning problems that are commonly encountered in machine
learning. The choice of which problem to use depends on the nature of the data and the task at hand.

4. Common machine learning problems faced while analyzing customer data and improving
customer engagement.

When analyzing customer data and improving customer engagement, there are several common
machine-learning problems that can be encountered. Here are some examples:

 Lack of quality data: Ensuring the quality of customer data can be a significant hurdle to
overcome. It is common to encounter issues such as incomplete, inconsistent, or inaccurate data,
which can lead to the creation of noisy predictive models or erroneous conclusions. These data
quality challenges pose a risk to the reliability and effectiveness of analytical processes. In order to
mitigate these issues, it is crucial to implement robust data cleansing, validation, and verification
procedures. By addressing data completeness, resolving inconsistencies, and rectifying
inaccuracies, organizations can improve the reliability of their predictive models and draw more
accurate conclusions from their data.
 Lack of labeled data: Most machine learning algorithms require labeled data to train a model and
make predictions. However, capturing a labeled data set can be costly. For instance, in customer
data, there might be a lack of labeled data representing what messages, ads, promotions, etc.
were really successful in gaining engagement or conversion.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Over fitting: Over fitting is common when a machine learning model is too complex or tries to fit
too closely to the training data. This can result in a model that doesn't generalize well and makes
poor predictions for new and unseen data.

 Interpretability: Many machine learning models are black boxes and provide no interpretation for
how they make decisions. This can be troublesome when designing models for improving
customer engagement. It is often necessary to know why a model makes certain predictions.

 Poor performance: Predictive models can sometimes lack the desired level of performance due to
inappropriate choice of algorithms, features, or hyperparameters. They may also underperform
because of the data quality, reusability, or nature of the problem.

 Incorporating the human element: Sometimes, improving customer engagement requires human
touch or understanding, and machine learning models cannot fully solve the problem. There may
be non-machine learning, human efforts, and methods, which can provide insights outside the
machine learning.
These problems can be tackled by fine-tuning the chosen machine learning model on the data set;
improving the quality and quantity of the data, testing algorithms with benchmark models, and
augmenting it with business rules, as their combination is more powerful. Choosing the right
algorithm, enhancing data quality, testing, and fine-tuning can also help address them.

P2. “Demonstrate the taxonomy of machine learning algorithms.”

5. Taxonomy of machine learning:

The taxonomy of machine learning algorithms refers to the classification of different machine learning
techniques based on their functionality and the ways they process and learn from data. It is a way to
categorize and organize machine learning models based on their inputs, processes, and outputs.
There are commonly three main categories of machine learning algorithms based on their

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
functionality: supervised learning, unsupervised learning, and reinforcement learning. These
categories can be further divided into specific subcategories based on the type of learning problem or
technique used, such as clustering, neural networks, decision trees, and more. Here's a brief
overview of each category,

1. Supervised Learning:

Supervised learning algorithms are used to train a model on labeled data. Supervised learning is a
type of machine learning where the model is trained on labeled data, wherein there are input
variables and corresponding output variables. The goal is to allow the model to learn the relationship
between the input and output data to make accurate predictions. Each instance of the input data has
a corresponding target output. Supervised learning algorithms learn a function that can map new
input data to its target output. The supervised learning process involves preparing the datasets,
selecting the correct algorithm, and training and testing the model on the data. During the training
phase, the model adjusts its weights and biases to fit the input data and correct output labels. Once
the model is trained, it is tested on new data, and its accuracy is determined by comparing the
predicted outputs with the actual outputs. Supervised learning can be used for both classification and
regression tasks. In classification, the model learns to predict the categorical label or class of the input
data, while in regression, the model learns a function that maps the input data to a continuous output
variable. Some of the most commonly used supervised learning algorithms include Linear Regression,
Logistic Regression, Decision Trees, Random Forests, and Support Vector Machines (SVM).

Supervised learning is widely used in various fields such as image and speech recognition, natural
language processing, sentiment analysis, and fraud detection.

Some common types of supervised learning algorithms are;

 Linear Regression:
Linear regression is a regression-based supervised learning algorithm used to predict a continuous
output value based on single or multiple input variables. The algorithm creates a linear relationship
between the input variables and the output variable.

 Logistic Regression:
Logistic regression is a classification-based supervised learning algorithm used to predict the
probability of a categorical variable or class based on one or more input variables. The algorithm maps
the input variables to a probability distribution over the output variable.
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Decision Trees:
Decision Trees are a classification and regression-based supervised learning algorithm used to predict
the value of a target variable based on multiple input variables. The algorithm splits the data based on
successive questions, forming a tree-like structure that helps classify the data.

 Random Forests:
Random Forests is a classification and regression-based supervised learning algorithm that works by
creating multiple decision trees and combining their outputs to create a final prediction. Each tree in
the forest is built on a random subset of features and data to prevent overfitting.

 Support Vector Machines (SVM):

Support Vector Machines are a classification-based supervised learning algorithm used to find the
hyperplane that best separates the input data based on the input features. This algorithm works well
with high-dimensional data and is particularly suited for binary classification problems.

Overall, supervised learning algorithms play an essential role in most machine learning applications,
including image and speech recognition, sentiment analysis, fraud detection, and more. By
understanding and applying these algorithms, machine learning models can make accurate
predictions, improve efficiency, and help automate decision-making processes.

2. Unsupervised Learning:

Unsupervised learning algorithms are used when there is no target output for the input data, which
means the input data is not labeled. It is used when data does not have any label, making it
impossible for the model to learn to predict outputs. The goal of unsupervised learning is to identify
patterns or structures in the data and to identify underlying patterns and relationships that may not
be visible to the human eye. Unsupervised learning algorithms are useful for exploring large datasets
and detecting hidden patterns and relationships that can be used to gain insights or improve other
processes. These algorithms also help in identifying data quality issues, such as missing values or
outliers, by plotting the data in lower-dimensional spaces. Some common types of unsupervised
learning algorithms are:

 Clustering Algorithms:
Clustering is an unsupervised learning algorithm that involves grouping similar data points together
based on their similarities, using algorithms like k-means clustering or hierarchical clustering. It’s
a group of similar data points together based on their similarity grouping customers based on their
purchasing behaviour or clustering images based on their features.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Dimensionality Reduction Algorithms:
Are used when there are a large number of input features, which can be used to reduce the
complexity of the problem.

 Association Rule Learning Algorithms:

Association rule mining is an unsupervised learning method used to extract hidden patterns or
relationships in large datasets. This algorithm scans through large datasets to find associations among
the variables and generate conditional probability rules. It is to identify relationships between
variables in large datasets.

 Autoencoders:
The autoencoder is an unsupervised learning algorithm that learns a representation for the input data
by minimizing the difference between the input and output data points. It is used for feature
extraction, dimensionality reduction, and anomaly detection.

 Principal Components Analysis (PCA):

Principal Components Analysis is an unsupervised learning technique that is used for dimensionality
reduction. It transforms high-dimensional, correlated variables into a smaller set of uncorrelated
variables called principal components.

3. Reinforcement Learning:

Reinforcement learning algorithms are a type of learning in which the algorithm learns by interacting
with the environment to produce an output. The algorithm receives feedback or rewards based on its
output and learns to optimize its decision-making process. Reinforcement learning involves an agent
interacting with its environment by taking actions and receiving feedback in the form of rewards or
punishments based on its actions. These rewards incentivize the agent to learn and optimize its
behavior in response to the environment. Reinforcement learning algorithms are used to solve
sequential decision-making problems.

The main elements of reinforcement learning are:

State: The current state of the environment.

Action: The decision or action chosen by the agent in response to the state.

Reward: The feedback received by the agent in response to its action.

Policy: The strategy or rule used by the agent to determine its actions based on the current state.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Common types of reinforcement learning algorithms:

There are several types of reinforcement learning algorithms. Here are some of the most common
types:

 Q-Learning:
Q-Learning is a value-based reinforcement learning algorithm that learns the optimal action selection
strategy using a quality function called the Q function. The Q function represents the expected
rewards of an agent when choosing an action in a given state.

 Deep Q-Networks (DQN):

Deep Q Networks (DQN) is an extension of Q-learning that uses deep neural networks to approximate
the Q function. This algorithm enables agents to interact with high-dimensional environments and can
learn from raw sensory inputs, such as pixels.

 Policy Gradients:
Policy Gradients are a policy-based reinforcement learning algorithm that uses gradient descent to
optimize the policy function. The policy function maps the state of the environment to the action that
should be taken in that state.

 Actor-Critic:
Actor-Critic is a hybrid of both policy-based and value-based reinforcement learning algorithms that
combines the best of both worlds. The Actor chooses an action and an associated value function
called the Critic that estimates how good the choice was being used to update the actor parameters.

 Monte-Carlo Tree Search (MCTS):

Monte-Carlo Tree Search (MCTS) is a search-based reinforcement learning algorithm that is designed
for decision-making problems in which the full state cannot be observed. It performs a tree search on
a discrete decision space and often uses randomness to guide its decisions.

These algorithms are commonly used in various applications such as robotics, game playing,
recommender systems, and recommendation engines. By learning the optimal policies or strategies
needed to perform specific tasks, these algorithms improve automatically the efficiency and
effectiveness of decision-making processes in many industries.

Difference between supervised Learning, Unsupervised Learning and Reinforcement Learning:

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Supervised Learning Unsupervised Learning Reinforcement
Learning

Data Labeled data Unlabeled data Feedback-driven

interactions with the
environment

Learning goal Predictive modeling, Discovering patterns, Learning optimal

classification, relationships, policies for decisions-
structures making
regression

Feedback Mechanism Explicit feedback No explicit feedback Feedback through

through labeled data rewards or penalties

Training Process Learning from labeled Learning from Learning through trial
data unlabeled data and error

Typical Algorithms Decisions trees, SVM, Clustering, Q-learning, Policy

Neural Networks Dimensionally Gradients Methods
Reduction

Application Examples Email spam, Clustering Autonomous driving,

classification, image segmentation, game playing
recognition anomaly detection

P3. Investigate a range of machine learning algorithms and how these algorithms solve the learning
problems.

Machine learning algorithms are a crucial component of artificial intelligence systems. They enable
machines to learn from data and make predictions or decisions without being explicitly programmed.
There are various types of machine learning algorithms, each designed to solve different types of
learning problems. We’ll look at some of the commonly used algorithms and how they address these
problems.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
1. Regression Algorithms:

Regression algorithms are a type of supervised learning algorithm used to predict the continuous
output variable from a set of input variables. Regression algorithms learn a function that maps the
input variables to the output variable, and this function can be used to make predictions on new and
unseen data. Here are some popular Regression algorithms:

 Linear Regression:
Linear Regression is a commonly used regression algorithm that learns a linear relationship between
the input variables and the output variable. It is used to predict the continuous output variable where
the relationship between the input and output variables is linear.

 Polynomial Regression:
Polynomial Regression is a type of regression algorithm used to capture non-linear relationships
between the input variables and the output variable. It involves fitting a polynomial function to the
input data to approximate the output variable.

 Ridge Regression:
Ridge Regression is a regularized linear regression algorithm that adds a penalty term to the loss
function. This penalty term helps to reduce the impact of noisy features and prevents overfitting in
high-dimensional data.

 Lasso Regression:
Lasso Regression is another regularized linear regression algorithm similar to Ridge Regression but
uses a different penalty term. Lasso Regression tends to produce sparse models by setting some of
the coefficients to zero and selecting only the most significant features.

 Decision Trees Regression:

Decision Trees Regression is a non-parametric regression algorithm that involves recursively
partitioning the input space into regions with homogeneous output values. The output value of a new
data point is predicted by its location in these regions.

 Random Forest Regression:

Random Forest Regression is an ensemble-based regression algorithm that involves combining
multiple decision trees trained on different subsets of the input data. It reduces the variance of the
model and improves the accuracy of the predictions.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Overall, Regression algorithms are widely used in various applications such as finance, economics, and
healthcare. Their ability to predict the continuous output variable makes them particularly useful
when dealing with real-valued data. By understanding the strengths and weaknesses of different
regression algorithms, data scientists can choose the right model to improve the accuracy and
efficiency of their predictions.

2. Classification Algorithms:

Classification algorithms are a kind of machine learning algorithm that is used to separate input data
into different classes or categories based on their features. These algorithms are trained on labeled
data, which means the class label of each data point is known in advance. Once trained, classification
algorithms can be used to predict the class label of new, unseen data.

Here are some popular classification algorithms:

 Logistic Regression:

Logistic Regression is a linear classification algorithm that uses a logistic function to model the
probability of an input variable belonging to a certain class. It is widely used for binary classification
problems where the output variable has two possible outcomes.

 Decision Trees:
Decision Trees are a nonlinear classification algorithm that learns a tree-like structure by
systematically splitting the input data based on successive questions. It is useful in multiclass
classification problems where there are more than two possible outcomes.

 Random Forests:
Random Forests builds an ensemble of decision trees by randomly selecting a subset of the input data
and a subset of the features to train each tree. It is used to reduce the variance of the model and
improve the accuracy of the predictions.

 Naive Bayes:
Naive Bayes is a probabilistic classification algorithm that uses Bayes' theorem to predict the class
label of an input variable. It assumes that the input variables are independent of each other, which
makes it a computationally efficient algorithm.

 Support Vector Machines (SVM):

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Support Vector Machines is a linear or nonlinear classification algorithm that finds the hyperplane
that best separates the input data into classes. It is useful in high-dimensional data and can be
extended to multiclass classification problems.

These algorithms are widely used in various fields such as image and speech recognition, sentiment
analysis, fraud detection, and many more. By choosing the right classification algorithm, data
scientists can improve their models' accuracy and efficiency, leading to better decision-making
processes.

3. Clustering Algorithms:

Clustering algorithms are an unsupervised machine learning technique that groups similar data points
together based on their similarities without any prior knowledge or labeled data. The goal of
clustering is to identify patterns or structures in the data to help understand the underlying
relationships between variables. Here are some popular clustering algorithms:

 K-means Clustering:
K-means clustering is a popular clustering algorithm that partitions the input data into k clusters
based on the distance between the data points. The algorithm works by iteratively assigning data
points to the nearest cluster center and updating the center of each cluster. It can handle large data
sets and is easy to implement.

 Hierarchical Clustering:
Hierarchical clustering is a clustering algorithm that builds a hierarchy of groups by recursively
splitting data points into smaller clusters. The algorithm works by merging or splitting clusters based
on a distance metric or similarity measure. It can be visualized as a dendrogram and is useful for
exploring the structure of the data.

 Density-Based Spatial Clustering of Applications (DBSCAN):

DBSCAN is a clustering algorithm that groups data points together based on their density. The
algorithm works by identifying core points that have a minimum number of neigh boring points and
connecting them to nearby points to form a cluster. It can handle noise, outliers, and irregularly
shaped clusters.

 Mean-Shift Clustering:

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Mean-shift clustering is a clustering algorithm that involves shifting the data points toward the center
of the nearest cluster until they converge to a stable configuration. The algorithm works by identifying
and updating the centers of the clusters based on the density of the data points. It can handle
multiple modes or clusters.

 Expectation-Maximization (EM) Clustering:

EM clustering is a probabilistic clustering algorithm that involves finding the maximum likelihood
estimate of the parameters in a mixture model of the data. The algorithm works by iteratively
estimating the hidden parameters that generate the data and assigning the data points to the most
probable cluster. It can handle missing data and is useful in identifying sub-populations.

Clustering algorithms are used in various applications such as image segmentation, customer
segmentation, anomaly detection, and many more. By grouping similar data points together,
clustering algorithms help uncover useful patterns or information in the dataset.

4. Association Rule Learning Algorithms:

Association rule learning algorithms are a type of unsupervised machine learning technique used to
discover interesting relationships between variables in large datasets. They uncover hidden patterns
or relationships between variables that may not be apparent with simple data analysis techniques.
The algorithms are commonly used in business intelligence, e-commerce, web analytics, and many
more. Association rule learning algorithms are used in various applications such as market basket
analysis, website usage analysis, and customer transaction analysis to generate insights that help in
marketing, product and inventory management, and many more. By identifying these hidden patterns
or relationships between variables, association rule learning can help in generating insights that
enable organizations to enhance their business process and decision-making.

Here are some popular associated rule-learning algorithms:

 Apriori Algorithm:
Apriori Algorithm is a frequent itemset mining algorithm used to generate rules that show the
association between the items. It works by scanning the input data and identifying the frequent
itemset. The algorithm then generates the association rules from the frequent itemset based on a
specified minimum support and confidence threshold.

 FP-Growth Algorithm:
The frequent Pattern Growth (FP-Growth) Algorithm is an efficient algorithm used to mine frequent
item sets and association rules in large datasets. It works by building a frequent pattern growth tree,
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
which is a compact representation of the dataset. The FP-Tree can be used to generate frequent item
sets and association rules with high accuracy.

 Eclat (Equivalence Class Transformation) Algorithm:

Eclat Algorithm is a frequent itemset mining algorithm used to generate association rules from
transaction databases. It works by computing the intersection of frequent item sets, which is a more
efficient approach since it avoids generating the subset of an item set. The algorithm is commonly
used in market-basket analysis for finding items that are frequently purchased together.

4. Deep Learning Algorithms:

Deep learning algorithms are a type of machine learning technique that uses artificial neural networks
with multiple layers to learn hierarchical representations of input data. The algorithms use these
hierarchical representations to solve complex and large-scale machine-learning problems such as
image and speech recognition, natural language processing, and robotics. Deep learning algorithms
work by mimicking the structure and function of the human brain. They consist of multiple layers of
artificial neurons and learn by adjusting the weights and biases of these neurons through
backpropagation. Deep learning algorithms have significantly improved the state-of-the-art for
various machine learning problems. They are widely used in applications such as image and speech
recognition, natural language processing, autonomous driving, and many more. By learning complex
hierarchical representations of data, deep learning algorithms can detect subtle patterns and
relationships that may not be apparent with other machine learning algorithms. Here are some
popular deep-learning algorithms:

 Convolutional Neural Networks (CNNs):

Convolutional neural networks are a deep learning architecture used for image and video processing
tasks. They use convolutional layers to extract features from the input image and output the class
probabilities or predictions.

 Recurrent Neural Networks (RNNs):

Recurrent neural networks are a deep learning architecture used for sequential data processing such
as speech recognition and text processing. They use feedback loops with feedback from previous time
steps to preserve temporal dependencies and memory to process the sequence of inputs.

 Generative Adversarial Networks (GANs):

Generative adversarial networks are a deep learning architecture used for unsupervised learning tasks
such as image and video generation, and data augmentation. GANs use two neural networks - a

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
generator network that learns to generate samples similar to the input data and a discriminator
network that learns to distinguish between the generated samples and the real data.

 Autoencoders:
Autoencoders are a deep learning architecture used for learning a compact representation of the
input data by encoding the input data into a smaller representation and decoding it back to its original
form. They are used for data compression, demonising, and feature learning.

P4. Demonstrate the efficiency of these algorithms by implementing them using an appropriate
programming language or machine learning tool.
Machine learning methods are often implemented and demonstrated using a programming language
or machine learning tool. In this section, I'll show you how to implement and assess the efficiency of
an algorithm for supervised learning utilising Python and sci-kit-learn, an established machine learning
toolkit, as well as other libraries and functions as needed.

Python was the best programming language choice for this prediction system. Its line-by-line
execution is ideal for this type of project, and the extensive library set provided makes it the first
choice for most ML as well as AI developers. NumPy, pandas, seaborn, and sci-kit-plot libraries will be
used to improve the prediction model for this project. Python's capacity to manage and transform
data is critical for machine learning, which requires constant processing, and Google Collaboratory will
be utilized as the IDE. Furthermore, exploratory data analysis (EDA) will be performed to provide the
best possible outcomes, utilizing the tremendous features and benefits of Python outlined above.

Consider the following scenario: we want to forecast customer attrition based on the available data in
order to boost customer engagement. The code below employs the necessary libraries, data loading,
visualization, one-hot encoding, outlier or anomaly elimination, and all other procedures required to
train the model.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Python is often regarded as the best language for implementing and analysing machine learning
algorithms due to its broad library support and ease of use. Its powerful ecosystem includes well-
known libraries such as sci-kit-learn, NumPy, pandas, seaborn, and sci-kit-plot, which offer a wide
variety of data loading, pre-processing, visualisation, and model training capabilities. When it
comes to tasks like predicting customer turnover, Python's machine learning modules provide a
diverse set of supervised learning models to pick from. Logistic Regression, Decision Tree
Classifier, Random Forest Classifier, and ensemble approaches such as AdaBoost are just a few
examples of sophisticated algorithms that are easily implemented. These algorithms can
successfully capture patterns and linkages in data to anticipate customer churn, allowing firms to
engage at-risk clients proactively. Furthermore, Python's environment provides unsupervised
learning techniques, such as Hierarchical Clustering, that can be used to uncover hidden patterns
and segmentation in consumer data. Businesses can acquire insights into various client profiles
and modify their retention tactics accordingly by grouping comparable consumers together.
Businesses may improve customer engagement and predictive accuracy in tasks such as customer
churn prediction by combining Python's adaptability with a varied collection of machine learning
methods. The Python ecosystem's ability to readily integrate diverse libraries and tools makes it a
popular choice among data scientists and machine learning practitioners.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
M1. Evaluate the category of machine learning algorithms with appropriate examples.
Machine learning algorithms can be used in various industries such as healthcare, finance,
entertainment, and many more. By choosing the appropriate machine learning algorithm based on
the type of problem, data scientists can analyze complex data and extract valuable insights.

The category of machine learning algorithms can be broadly classified into three main categories:
Supervised Learning, Unsupervised Learning, and Reinforcement Learning.

1. Supervised Learning: Supervised learning algorithms are used to predict the values of an outcome
variable based on the input variables. In supervised learning, the algorithm is trained on labeled data
to learn the relationship between the input variables and output variables. Examples of supervised
learning algorithms include:

 Linear Regression: Used to predict continuous values such as predicting the price of a house based
on its features.
 Logistic Regression: Used to predict a binary outcome such as a credit risk score.
Decision Trees: Used in classification problems such as classifying customers into high-risk and low-
risk groups.

 Support Vector Machines: Used in classification problems such as email spam filtering
Neural Networks: Used in various applications such as image and speech recognition, natural
language processing, and many more.

2. Unsupervised Learning: Unsupervised learning algorithms are used to discover the underlying
structure of the input data where the output variable is unknown. Unsupervised learning is used to
find relationships and patterns in data. Examples of unsupervised learning algorithms include:

 Clustering: Used to identify similar groups of data points in a dataset.

 Principal Component Analysis (PCA): Used to reduce the dimensionality of a dataset.
 Association Rule Learning: Used in market basket analysis to identify which products are often
bought together.
 Autoencoders: Used in recommendation systems to generate recommendations based on user
behaviour or preferences.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
3. Reinforcement Learning: Reinforcement learning algorithms are used to train an agent to make
decisions by learning from feedback in a real or simulated environment. Examples of reinforcement
learning algorithms include:

 Q-Learning: Used to learn a policy that maximizes the expected reward in a Markov Decision
Process.
 Deep Q Networks: Used in various applications such as playing video games, robotics, and
autonomous driving.
 Monte Carlo Tree Search: Used in game-playing algorithms such as AlphaGo.
M2. Analyze these algorithms using an appropriate example to determine their power.
Evaluating machine learning algorithms involves assessing their performance and efficiency in solving
specific problems. The selection of an algorithm relies on factors such as the nature of the data, the
task at hand, and the desired outcomes. To comprehensively evaluate different algorithms, it is
essential to consider their accuracy, speed, and generalization capabilities to new data.

When it comes to analyzing customer data and improving customer engagement strategies,
unsupervised learning algorithms like k-means clustering can be highly valuable. By employing k-
means clustering, businesses can identify distinct patterns and segments within their customer base,
leading to more personalized and targeted approaches.

With k-means clustering, the algorithm groups customers based on similarities in their attributes or
behavior. This can uncover valuable insights, such as identifying clusters of customers with similar
preferences, purchase histories, or browsing behavior. By understanding these patterns, businesses
can tailor their engagement strategies, offering personalized recommendations, targeted marketing
campaigns, or loyalty programs, ultimately leading to improved customer satisfaction and retention.

The power of unsupervised learning algorithms like k-means clustering lies in their ability to uncover
hidden structures and patterns in customer data without relying on predefined labels. This flexibility
allows businesses to gain a deeper understanding of their customers and make data-driven decisions
to enhance customer engagement.

In conclusion, evaluating machine learning algorithms involves considering their performance metrics,
such as accuracy and speed, as well as their ability to generalize to new data. Unsupervised learning
algorithms like k-means clustering can be instrumental in analyzing customer data, identifying
patterns, and improving customer engagement strategies through personalized and targeted
approaches.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
1. Unsupervised learning:

The purpose of unsupervised learning is to extract meaningful patterns, structures, or relationships

from unlabeled data. Unsupervised learning algorithms, unlike supervised learning, do not have
access to explicit labels or target variables during training. Instead, these algorithms concentrate on
detecting intrinsic patterns or similarities in the data. Unsupervised learning's fundamental goal is to
investigate and comprehend the underlying structure of data without prior knowledge or assistance.
Clustering, dimensionality reduction, and anomaly detection are examples of such tasks.

Dimensionality reduction:

Dimensionality reduction approaches, such as principal component analysis (PCA) or t-SNE (t-
distributed stochastic neighbor embedding), seek to minimize the number of variables or dimensions
in a dataset while retaining critical information. This can be useful for visualizing large amounts of
data or isolating the most significant aspects for further research.

2. Anomaly Detection:

Anomaly detection methods find instances in data that are uncommon or abnormal that differ
significantly from the majority of observations. This can help detect fraudulent transactions, network
intrusions, and manufacturing flaws.

Unsupervised learning is essential in exploratory data analysis because it uncovers hidden patterns
and insights, facilitates data preparation, and serves as a foundation for later analysis or decision-
making.

It is crucial to highlight that evaluating unsupervised learning algorithms is often more difficult than
supervised learning algorithms because there are no explicit labels to compare against. The quality
and coherence of the revealed patterns or structures are frequently assessed using domain
knowledge or intrinsic criteria related to the job at hand.

3. Reinforcement Learning:

Reinforcement learning is a subset of machine learning that focuses on making decisions in dynamic
contexts. It is based on how humans and animals learn to maximize rewards or outcomes via trial and
error. An agent learns to execute behavior in an environment to maximize cumulative rewards over
time through reinforcement learning.

Primary Components:

The following are the primary components of reinforcement learning:

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
  Agent: The creature or system that interacts with the environment, performs actions, and
learns from feedback is referred to as an agent.

 Environment: The external system or context in which the agent functions is referred to as the
agent's environment. It can be a virtual world, a physical world, or even a software application.

 Actions: The many options or decisions available to the agent in the environment.

 Rewards: The feedback mechanism that tells the agent about the attractiveness of its
behaviours is known as rewards. Positive, negative, or zero rewards indicate the consequence
of an action.

The purpose of reinforcement learning is to discover an optimal policy—a state-to-action mapping

that maximises the expected cumulative rewards over time. The agent learns through an iterative
method that frequently includes the following steps:

 Action Selection: The agent chooses an action to take based on the observed state.

 Interaction with the Environment: The agent executes the chosen action in the environment,
which transitions to a new state.

 Reward and Feedback: The environment rewards or punishes the agent based on its actions,
offering feedback on the action's desirability.

 Learning and Updating: Based on the received incentive, the agent updates its knowledge or
policy in order to better future decision-making.

Reinforcement learning algorithms use a variety of ways to learn and improve the agent's policy,
including value-based methods (e.g., Q-learning), policy-based methods (e.g., REINFORCE), and actor-
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
critic methods (which combine value-based and policy-based approaches).Reinforcement learning
algorithms use a variety of ways to learn and improve the agent's policy, including value-based
methods (e.g., Q-learning), policy-based methods (e.g., REINFORCE), and actor-critic methods (which
combine value-based and policy-based approaches).aches). However, reinforcement learning can be
difficult to implement since it frequently necessitates large computer resources, precise reward
design, and balancing the exploration-exploitation trade-off in order to identify an ideal strategy.
Overall, reinforcement learning provides a powerful framework for learning decision-making in
dynamic contexts and has the potential to revolutionize several fields by enabling intelligent, adaptive
systems.

4. Q-learning:

Q-learning is a common model-free reinforcement learning algorithm that seeks to learn the best
policy for making decisions in a given environment. It is a value-based strategy for guiding the agent's
behaviours by learning the values of state-action pairings known as Q-values. The Q-learning method
continuously works through episodes of agent-environment interactions. A high-level overview of the
Q-learning process follows:

 Initialization: For each state-action pair, set the Q-values arbitrarily.

 Exploration and exploitation: The agent chooses an action to do based on the current situation at
each time step. The agent can use an exploration-exploitation approach, such as epsilon-greedy, in
which the agent either explores random actions with a specific probability (exploration) or
chooses the action with the highest Q-value (exploitation).
 Interaction with the Environment: The agent executes the chosen action and transitions to a new
state, earning a reward from the environment.
 Q-value Update: Once the reward has been received, the agent changes the Q-value of the
previous state-action combination based on the observed reward and the estimated future
rewards. The following is the Q-value update equation for Q-learning:
Q (s, a) = Q (s, a) + α * (r + γ * max (Q (s', a')) - Q (s, a))

In this equation, Q (s, a) represents the Q-value for a state-action pair (s, a), r is the observed
reward, s' is the new state, is the learning rate (0 to 1) that determines the weight given to new
information, and is the discount factor (0 to 1) that balances the importance of immediate and
future rewards.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Steps 2-4 are repeated iteratively: the agent continues to interact with the environment,
selecting actions, updating Q-values, and refining its policy.

The Q-values steadily converge towards their optimal values over numerous episodes, representing
the predicted cumulative rewards for each state-action pair. When the learning process is over, the
agent can implement the policy by selecting actions with the greatest Q-values to make judgements in
the environment. Q-learning is renowned for its capacity to manage complex settings with substantial
state and action spaces, which makes it appropriate for use in applications like game play and control
systems. The environment must have the Markov property in order for the future state to solely be
dependent on the present state and action, which may not always be the case.

To address problems and boost the algorithm's performance in various situations, extensions and
variations of Q-learning, such as Double Q-learning, Deep Q-Networks (DQN), and Duelling DQN, have
been introduced.

Q-learning is a fundamental reinforcement learning method that, by iteratively updating the Q-values
based on observed rewards, enables agents to acquire the best practises for making decisions in
dynamic situations.

D1. Critically evaluate why machine learning is essential to the design of intelligent machines
Machine learning is a field of artificial intelligence (AI) that allows machines to learn from data and
make predictions or decisions without being explicitly programmed. In essence, machine learning
allows the software to automatically self-improve and gain insight from data so that it can carry out
intelligent tasks and identify patterns from data sets. Machine learning algorithms are designed to
find patterns in data and identify relationships between them. The algorithms use these relationships
to make predictions or decisions about new data sets. Machine learning is useful in a wide range of
applications, such as image recognition, speech recognition, natural language processing, predictive
analytics, and fraud detection in finance. It is essential to the design of intelligent machines because it
enables them to learn and make predictions based on data, without being explicitly programmed. This
allows machines to adapt to different situations and make decisions based on experience, similar to
how humans learn and make decisions.

 Ability to learn and adapt:

Machine learning algorithms allow machines to learn from experience and adapt to new
situations. This is important because traditional programming relies on rules and requires specific

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 instructions for every possible scenario. Machine learning, on the other hand, can learn from
examples and patterns, and make predictions based on that information. This helps machines to
function more autonomously and handle more complex tasks.
 Improved accuracy:
Machine learning algorithms can identify and learn patterns within data, allowing machines to
make predictions and decisions with improved accuracy. This is because the algorithms are
designed to identify relationships between different data points and can use this information to
make predictions and decisions that might not be obvious otherwise.
 Efficient decision-making:
Machine learning algorithms can process vast amounts of data in real time and make decisions
based on that data. This is important because decisions made in real time and based on current
data can be more accurate and relevant than those made using static or general rules. This
enables machines to act more efficiently and react quickly to new situations.
 Enhanced capabilities:
Machine learning algorithms can enable machines to perform tasks that would be difficult or
impossible for humans to do manually. For example, image recognition, speech recognition, and
natural language processing all require complex data analysis and pattern recognition, which can
be difficult or impossible for humans to do without error. With machine learning, machines can
learn and adapt to these tasks and perform them with greater accuracy and efficiency.
 Continuous improvement:
Machine learning algorithms can continually improve over time as new data is added to the
system. This allows machines to adapt and improve their performance over time, becoming more
intelligent and capable. This is important because it means that machines can learn from their
mistakes, anticipate new situations, and continually adapt to changing requirements.
 Adaptive and autonomous functioning of machines:
Machine learning enables machines to become more adaptive and autonomous in their
functioning, making them more efficient and accurate in completing tasks.
 Allows for predictive modeling in various fields such as finance and healthcare:
Machine learning enables predictive modeling, which helps decision-makers in various fields to
make data-driven decisions based on predictions and trends.
 Enables automation of tasks, leading to increased efficiency and productivity:
Machine learning algorithms can automate tasks, which leads to increased efficiency and
productivity by reducing the manual workload for humans.
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
 Helps in fraud detection and cybersecurity:
Machine learning algorithms can identify patterns and anomalies in data, making it easier to
detect fraud and improve cyber security measures.
 Facilitates personalized recommendations for users based on their preferences and behavior:
Machine learning algorithms can analyze user data to provide personalized recommendations
based on their behaviour and past interactions.
 Can identify and classify vast amounts of data and detect patterns:
Machine learning algorithms can classify and identify data, making it easier to spot patterns that
could be valuable in making predictions and decisions.
 Helps in logistics and supply chain management through predictive analytics:
Machine learning algorithms can make predictions about future events or trends, making them
useful in logistics and supply chain management by enabling better planning and decision-making.
 Plays a critical role in information retrieval and search engine optimization:
Machine learning algorithms enable search engines to retrieve relevant information for search
queries, making it easier for users to find what they are looking for and search engines to retrieve
relevant information for search queries, making it easier for users.

 Helps in customer service and support with chatbots and virtual assistants:
By using machine learning to develop chatbots and virtual assistants, customers can receive
support and assistance without human intervention, leading to improved customer experience.
Machine learning algorithms have enabled the development of chatbots and virtual assistants that
can provide customers with support and assistance, improving customer experience.

 Allows for better risk assessment and management in insurance and finance industries:
By using machine learning to predict and identify risks, the insurance and finance industries can
mitigate risks. Machine learning can analyze vast amounts of data to predict future events and
minimize financial loss. Machine learning algorithms can predict and identify risks, making it easier
for the insurance and finance industries to manage and mitigate such risks.

Overall, machine learning is essential to the design of intelligent machines because it enables them to
learn, adapt, and improve their performance over time. This enables machines to become more

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
autonomous, efficient, and capable, and to perform tasks that would otherwise be difficult or
impossible for humans to do.

6. References:

 TechTarget. (n.d.). Definition - machine learning (ML). SearchEnterpriseAI. Retrieved from

https://www.techtarget.com/searchenterpriseai/definition/machine-learning-ML. Accessed May
05, 2023

 MIT Sloan School of Management. (n.d.). Machine Learning Explained. Retrieved from
https://mitsloan.mit.edu/ideas-made-to-matter/machine-learning-explained. Accessed May 13,
2023

 Great Learning. (n.d.). What is Machine Learning? Retrieved from

https://www.mygreatlearning.com/blog/what-is-machine-learning/. Accessed May 11, 2023

 Smith, J. (2022). Machine Learning Algorithms. GitHub Repository. Retrieved from

https://github.com/johnsmith/machine-learning-algorithms

 Sutton, R. S., & Barto, A. G. (1998). Reinforcement Learning: An Introduction (No. UCB/EECS-98-
203). University of California, Berkeley.

Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23
Pearson BTEC HN RQF Assignment Feedback Form Template Academic Year 2022/23