Introduction to Machine learning:

Arthur Samuel, an early American leader in the field of computer gaming and artificial
intelligence, coined the term “Machine Learning” in 1959 while at IBM. He defined machine
learning as “the field of study that gives computers the ability to learn without being explicitly
programmed “. However, there is no universally accepted definition for machine learning.
Machine learning is a growing technology which enables computers to learn automatically from
past data. Machine learning uses various algorithms for building mathematical models and
making predictions using historical data or information. Currently, it is being used for various
tasks such as image recognition, speech recognition, email filtering, Facebook auto-tagging,
recommender system, and many more.
Machine Learning is said as a subset of artificial intelligence that is mainly concerned with the
development of algorithms which allow a computer to learn from the data and past
experiences on their own. The term machine learning was first introduced by Arthur Samuel in
1959.
We can define it in a summarized way as:
Machine learning enables a machine to automatically learn from data, improve performance
from experiences, and predict things without being explicitly programmed.
With the help of sample historical data, which is known as training data, machine learning
algorithms build a mathematical model that helps in making predictions or decisions without
being explicitly programmed. Machine learning brings computer science and statistics together
for creating predictive models. Machine learning constructs or uses the algorithms that learn
from historical data. The more we will provide the information, the higher will be the
performance.
A machine has the ability to learn if it can improve its performance by gaining more data.

Motivation for Machine Learning:

Machine learning is motivated by the desire to develop intelligent systems that can
automatically learn and improve from experience without explicit programming. Here are some
key motivations for machine learning:
Handling complex and large-scale data: With the exponential growth of data in various
domains, traditional rule-based or manually programmed approaches often struggle to extract
meaningful insights or make accurate predictions. Machine learning techniques enable the
analysis and interpretation of complex and large-scale datasets by automatically learning
patterns, relationships, and structures within the data.
Automation and efficiency: Machine learning algorithms can automate labor-intensive tasks
and decision-making processes that would otherwise require significant human effort and time.
By learning from historical data, models can automate repetitive tasks, make predictions,
classify objects, detect anomalies, and perform other complex tasks more efficiently and at
scale.
Adaptability and generalization: Machine learning models can adapt to changing conditions
and generalize knowledge to new, unseen data. Instead of relying on fixed rules or hard-coded
instructions, machine learning algorithms learn from examples and experience, allowing them
to make accurate predictions or decisions on novel inputs. This adaptability is particularly
valuable in dynamic environments or when dealing with complex, real-world problems.
Extracting valuable insights: Machine learning techniques can uncover hidden patterns,
correlations, and insights from data that might not be easily detectable by humans. By
analyzing vast amounts of data, machine learning models can identify trends, make predictions,
and generate actionable insights that can drive informed decision-making and improve various
processes across industries, such as healthcare, finance, marketing, and manufacturing.
Handling uncertainty and noisy data: Real-world data is often incomplete, noisy, or contains
inherent uncertainty. Machine learning algorithms are designed to handle such data and can
effectively deal with missing values, outliers, and noise. They can learn robust representations
and make informed decisions even in the presence of uncertainty, making them suitable for
applications where precise and reliable predictions are required.
Prediction and Decision-Making: Machine learning algorithms can analyze historical data and
learn patterns to make predictions about future events or outcomes. This predictive capability
is valuable in fields such as finance, healthcare, marketing, and weather forecasting, where
accurate predictions can inform critical decisions and improve outcomes.

Applications of Machine learning:

1. Image Recognition: Image recognition is one of the most common applications of machine
learning. It is used to identify objects, persons, places, digital images, etc.
2. Speech Recognition: While using Google, we get an option of "Search by voice," it comes
under speech recognition, and it's a popular application of machine learning. Speech
recognition is a process of converting voice instructions into text, and it is also known as
"Speech to text", or "Computer speech recognition." At present, machine learning algorithms
are widely used by various applications of speech recognition. Google assistant, Siri, Cortana,
and Alexa are using speech recognition technology to follow the voice instructions.
3. Traffic prediction: If we want to visit a new place, we take help of Google Maps, which shows
us the correct path with the shortest route and predicts the traffic conditions. It predicts the
traffic conditions such as whether traffic is cleared, slow-moving, or heavily congested with the
help of two ways:
o Real Time location of the vehicle form Google Map app and sensors
o Average time has taken on past days at the same time.
Everyone who is using Google Map is helping this app to make it better. It takes information
from the user and sends back to its database to improve the performance.
4. Product recommendations: Machine learning is widely used by various e-commerce and
entertainment companies such as Amazon, Netflix, etc., for product recommendation to the
user. Whenever we search for some product on Amazon, then we started getting an
advertisement for the same product while internet surfing on the same browser and this is
because of machine learning. Google understands the user interest using various machine
learning algorithms and suggests the product as per customer interest. As similar, when we use
Netflix, we find some recommendations for entertainment series, movies, etc., and this is also
done with the help of machine learning.
5. Self-driving cars: One of the most exciting applications of machine learning is self-driving
cars. Machine learning plays a significant role in self-driving cars. Tesla, the most popular car
manufacturing company is working on self-driving car. It is using unsupervised learning method
to train the car models to detect people and objects while driving.
6. Email Spam and Malware Filtering: Whenever we receive a new email, it is filtered
automatically as important, normal, and spam. We always receive an important mail in our
inbox with the important symbol and spam emails in our spam box, and the technology behind
this is Machine learning. Below are some spam filters used by Gmail:
o Content Filter
o Header filter
o General blacklists filter
o Rules-based filters
o Permission filters
Some machine learning algorithms such as Multi-Layer Perceptron, Decision tree, and Naïve
Bayes classifier are used for email spam filtering and malware detection.
7. Virtual Personal Assistant: We have various virtual personal assistants such as Google
assistant, Alexa, Cortana, Siri. As the name suggests, they help us in finding the information
using our voice instruction. These assistants can help us in various ways just by our voice
instructions such as Play music, call someone, open an email, Scheduling an appointment, etc.
These virtual assistants use machine learning algorithms as an important part. These assistant
record our voice instructions, send it over the server on a cloud, and decode it using ML
algorithms and act accordingly.
8. Online Fraud Detection: Machine learning is making our online transaction safe and secure
by detecting fraud transaction. Whenever we perform some online transaction, there may be
various ways that a fraudulent transaction can take place such as fake accounts, fake ids, and
steal money in the middle of a transaction. So to detect this, Feed Forward Neural network
helps us by checking whether it is a genuine transaction or a fraud transaction.
9.Smart Home Automation: Machine learning enables the automation of various tasks in smart
homes, such as adjusting lighting, temperature control, and predicting user preferences based
on historical data. It enhances convenience, energy efficiency, and personalized home
experiences.

Learning Association:
Learning associations in machine learning refers to the process of identifying relationships or
connections between different variables or features in a dataset. The goal is to learn and
understand how changes in one variable may relate to changes in another variable.
There are different techniques and algorithms in machine learning that can be used to learn
associations. Here are a few common approaches:
1)Association rule learning is a type of unsupervised learning technique that checks for the
dependency of one data item on another data item and maps accordingly so that it can be
more profitable. It tries to find some interesting relations or associations among the variables of
dataset. It is based on different rules to discover the interesting relations between variables in
the database.
The association rule learning is one of the very important concepts of machine learning, and it
is employed in Market Basket analysis, Web usage mining, continuous production, etc. Here
market basket analysis is a technique used by the various big retailer to discover the
associations between items. We can understand it by taking an example of a supermarket, as in
a supermarket, all products that are purchased together are put together. For example, if a
customer buys bread, he most likely can also buy butter, eggs, or milk, so these products are
stored within a shelf or mostly nearby. Consider the below diagram:
Association rule learning can be divided into three types of algorithms:
1. Apriori.
2. Eclat.
3. F-P Growth Algorithm.
Apriori Algorithm: This algorithm uses frequent datasets to generate association rules. It is
designed to work on the databases that contain transactions. This algorithm uses a breadth-first
search and Hash Tree to calculate the item set efficiently. It is mainly used for market basket
analysis and helps to understand the products that can be bought together. It can also be used
in the healthcare field to find drug reactions for patients.
Eclat Algorithm: Eclat algorithm stands for Equivalence Class Transformation. This algorithm
uses a depth-first search technique to find frequent item sets in a transaction database. It
performs faster execution than Apriori Algorithm.
F-P Growth Algorithm: The F-P growth algorithm stands for Frequent Pattern, and it is the
improved version of the Apriori Algorithm. It represents the database in the form of a tree
structure that is known as a frequent pattern or tree. The purpose of this frequent tree is to
extract the most frequent patterns.
How associate learning works:
Association rule learning works on the concept of If and Else Statement, such as if A then B.
Here the If element is called antecedent, and then statement is called as Consequent. These
types of relationships where we can find out some association or relation between two items is
known as single cardinality.These metrics are given below:
o Support
o Confidence
o Lift
Support: Support is the frequency of A or how frequently an item appears in the dataset.
{Percentage of transaction that contains both a and b}{A=>B = P(A∧B)}
Confidence: Confidence indicates how often the rule has been found to be true. Or how often
the items X and Y occur together in the dataset when the occurrence of X is already given.
Lift: It is the strength of any rule.
2) Correlation Analysis: Correlation analysis measures the statistical relationship between two
variables. It calculates a correlation coefficient that indicates the strength and direction of the
relationship. Positive correlation means that as one variable increases, the other tends to
increase, while negative correlation means they tend to change in opposite directions. Pearson
correlation coefficient and Spearman correlation coefficient are commonly used for measuring
correlations.
3)Decision Trees: Decision trees can capture associations between features by constructing a
tree-like structure that represents decisions or rules based on the input features. The splitting
criteria of decision trees are often based on association measures such as information gain or
Gini index
4)Neural Networks: Neural networks can learn complex associations by adjusting the weights
and connections between neurons during the training process. Deep learning architectures,
such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), are
commonly used to learn associations in image recognition, natural language processing, and
sequential data analysis tasks.

Classification of Machine Learning algorithms:

At a broad level, machine learning can be classified into three types:
1. Supervised learning
2. Unsupervised learning
3. Reinforcement learning
1) Supervised Learning: Supervised learning is a type of machine learning method in which we
provide sample labeled data to the machine learning system in order to train it, and on that
basis, it predicts the output.
The system creates a model using labeled data to understand the datasets and learn about
each data, once the training and processing are done then we test the model by providing a
sample data to check whether it is predicting the exact output or not.
The goal of supervised learning is to map input data with the output data. The supervised
learning is based on supervision, and it is the same as when a student learns things in the
supervision of the teacher. The example of supervised learning is spam filtering.
Supervised learning can be grouped further in two categories of algorithms:

Regression: Regression algorithms predict continuous numerical values. Examples include linear
regression, polynomial regression, and support vector regression.
Classification: Classification algorithms predict discrete class labels or categories. Examples
include logistic regression, decision trees, random forests, naive Bayes, support vector
machines (SVM), and artificial neural networks.
2) Unsupervised Learning: Unsupervised learning is a learning method in which a machine
learns without any supervision.
The training is provided to the machine with the set of data that has not been labeled,
classified, or categorized, and the algorithm needs to act on that data without any supervision.
The goal of unsupervised learning is to restructure the input data into new features or a group
of objects with similar patterns.
In unsupervised learning, we don't have a predetermined result. The machine tries to find
useful insights from the huge amount of data. It can be further classifieds into two categories of
algorithms:
Clustering: Clustering algorithms group similar data points together based on their similarities.
Examples include k-means clustering, hierarchical clustering, and density-based clustering (e.g.,
DBSCAN).
Association: Association algorithms are commonly used in unsupervised learning to discover
relationships, patterns, and associations among items or features within a dataset.
3) Reinforcement Learning: Reinforcement learning is a feedback-based learning method, in
which a learning agent gets a reward for each right action and gets a penalty for each wrong
action. The agent learns automatically with these feedbacks and improves its performance. In
reinforcement learning, the agent interacts with the environment and explores it. The goal of
an agent is to get the most reward points, and hence, it improves its performance.
The robotic dog, which automatically learns the movement of his arms, is an example of
Reinforcement learning
4) Semi-Supervised Learning Algorithms: These algorithms utilize a combination of labeled and
unlabeled data for learning. They leverage the limited labeled data to guide the learning
process while benefiting from the larger amount of unlabeled data to improve performance.
Examples include self-training, co-training, and multi-view learning.

Classification: Classification in machine learning is a supervised learning task that

involves categorizing or assigning data instances into predefined classes or categories based on
their features. The goal is to learn a model that can accurately classify new, unseen data based
on the patterns learned from labeled training data.
For instance, an algorithm can learn to predict whether a given email is spam or ham (no spam),
as illustrated below.

Before diving into the classification concept, we will first understand the difference between
the two types of learners in classification: lazy and eager learners. Then we will clarify the
misconception between classification and regression.
Lazy Learners Vs. Eager Learners
There are two types of learners in machine learning classification: lazy and eager learners.
Eager learners are machine learning algorithms that first build a model from the training
dataset before making any prediction on future datasets. They spend more time during the
training process because of their eagerness to have a better generalization during the training
from learning the weights, but they require less time to make predictions.
Lazy learners or instance-based learners, on the other hand, do not create any model
immediately from the training data, and this is where the lazy aspect comes from. They just
memorize the training data, and each time there is a need to make a prediction, they search for
the nearest neighbor from the whole training data, which makes them very slow during
prediction.
Here are the key components and steps involved in classification:
Dataset: In classification, we start with a set of data that is already labeled with categories. For
example, if we want to build a model to identify whether an email is spam or not, we need a
dataset where each email is already marked as spam or not spam.
Feature Extraction: We look at the data and try to identify the important characteristics or
features that can help us determine the category. For example, in the case of email spam
detection, features could include the presence of certain keywords, the length of the email, or
the frequency of certain words.
Model Selection: We choose a method or algorithm that will learn from the labeled data and
help us make predictions. There are different algorithms available, such as decision trees,
logistic regression, or support vector machines. Each algorithm has its own way of learning from
the data.
Model Training: We train the chosen model using our labeled data. The model looks at the
features of the data and tries to find patterns or relationships that can help it make accurate
predictions. It learns from the examples in the training data and adjusts its internal parameters
to improve its predictions.
Model Evaluation: Once the model is trained, we need to check how well it performs on new,
unseen data. We use a separate portion of our labeled data, called the test set, to evaluate the
model's accuracy. Evaluation metrics, such as accuracy (how often the model is correct) or
precision and recall (measuring how well the model identifies the positive and negative cases),
help us assess the model's performance.
Prediction: After the model is trained and evaluated, we can use it to make predictions on new,
unlabeled data. For example, we can use the trained model to predict whether an incoming
email is spam or not by analyzing its features and applying the learned patterns.
Model Optimization and Tuning: If the model's performance is not satisfactory, we can make
adjustments to improve it. We can try different settings or configurations, known as
hyperparameters, of the chosen algorithm. We can also experiment with different features or
data preprocessing techniques to enhance the model's accuracy.
Regression:
Regression in machine learning is a technique used to predict or estimate continuous numerical
values. It helps us understand the relationship between a dependent variable (the one we want
to predict) and one or more independent variables (the ones we use to make predictions).
Here's an explanation of each step of regression in easy language:
Dataset: You have a set of data that includes information about something you're interested in,
such as houses. For each house, you have the size (in square feet) and the price (in dollars).
Model Selection: You choose a method or approach that will help you predict the price of a
house based on its size. There are different methods available, but for now, let's focus on one
called linear regression.
Model Training: You use the data you have to train the model. The model looks at the
relationship between the sizes and prices of houses in the data and tries to find a straight line
that best fits the data points. This line represents a pattern that the model learns from the data.
Model Evaluation: Once the model is trained, you need to check how well it predicts the prices
of houses. You use some of the data that the model hasn't seen before to see how close its
predictions are to the actual prices. You measure the accuracy of the model using a metric called
mean squared error, which tells you how much the predicted prices differ from the actual prices
on average.
Prediction: Now that the model is trained and evaluated, you can use it to predict the price of a
new, unseen house based on its size. You provide the size of the house to the model, and it gives
you an estimated price based on the pattern it learned from the training data.
Model Optimization: If the model's predictions are not accurate enough, you can make
adjustments to improve it. You can try different approaches or techniques, such as using a more
complex model or adding more features to the data. The goal is to find the best-fitting line or
curve that captures the relationship between house sizes and prices.
Some types of regression are as follows:
Linear Regression: Linear regression is like drawing a straight line through data points. It finds
the best-fitting line that represents the relationship between the input variables (features) and
the output variable (target). It helps us make predictions based on a linear pattern.
Polynomial Regression: Polynomial regression is an extension of linear regression. It fits a curve
instead of a straight line to capture more complex relationships between the features and the
target variable.
Decision Tree Regression: Decision tree regression is like playing a guessing game. It asks
questions based on features and splits the data into different branches to predict the target
variable. It helps when the relationship between features and the target is non-linear and can
have complex interactions.
Random Forest Regression: Random forest regression is an ensemble method that combines
multiple decision trees. It averages the predictions from all the trees to make a final prediction.
It helps to reduce overfitting and improve the accuracy of predictions.

The origin of Machine Learning:

The origin of machine learning can be traced back to the development of artificial intelligence
(AI) as a field of study. The concept of machine learning emerged from the idea of creating
computer systems that can learn and adapt from data without being explicitly programmed.
Although the roots of AI can be traced back to early pioneers like Alan Turing and his work on
machine intelligence in the 1950s, the term "machine learning" itself was coined by Arthur
Samuel in 1959. Samuel, an American computer scientist, defined machine learning as the
ability of computers to learn from data and improve their performance over time without being
explicitly programmed.
In the early years, machine learning algorithms were primarily focused on symbolic approaches,
such as expert systems and rule-based reasoning. However, as computing power increased and
more data became available, researchers began exploring statistical and probabilistic methods
for machine learning.
The field of machine learning experienced significant advancements and breakthroughs in the
latter half of the 20th century and early 21st century.
Some notable milestones include:
The development of neural networks: In the 1950s and 1960s, researchers started exploring
artificial neural networks inspired by the structure and function of the human brain. The
perceptron algorithm, proposed by Frank Rosenblatt in 1958, laid the foundation for neural
network learning.
The rise of statistical learning: In the 1980s and 1990s, statistical learning approaches gained
prominence. Techniques such as linear regression, logistic regression, and decision trees
became popular for pattern recognition and prediction tasks.
Support Vector Machines (SVM): Introduced by Vladimir Vapnik and his colleagues in the
1990s, SVM is a powerful algorithm for classification and regression tasks. It works by finding an
optimal hyperplane that separates different classes in the feature space.
Deep learning revolution: Deep learning, a subfield of machine learning based on artificial
neural networks with multiple layers, experienced a resurgence in the early 2000s. Advances in
computational power, big data availability, and algorithmic innovations (e.g., convolutional
neural networks and recurrent neural networks) led to significant breakthroughs in areas like
computer vision, natural language processing, and speech recognition.

Uses and abuses of Machine learning:

1) Image recognition: The image recognition is one of the most common uses of machine
learning applications. The face recognition is also one of the great features that have
been developed by machine learning only. It helps to recognize the face and send the
notifications related to that to people.
2) Voice recognition: Machine learning (ML) also helps in developing the application for
voice recognition. It will help you to find the information when asked over the voice.
After your question, that assistant will look out for the data or the information that has
been asked by you and collect the required information to provide you with the best
answer. There are many devices available in today world of Machine learning for voice
recognition that is Amazon echo and googles home is the smart speakers.
3) Fraud Detection: Machine learning algorithms can detect patterns and anomalies in
data, helping identify fraudulent activities in areas like banking, insurance, and
cybersecurity.
4) Healthcare and Medicine: Machine learning is used for disease diagnosis, drug
discovery, medical image analysis, patient monitoring, and personalized medicine.
5) Autonomous Vehicles: Machine learning algorithms play a crucial role in self-driving
cars, enabling object detection, path planning, and decision-making based on real-time
sensor data.
6) Recommender Systems: Machine learning algorithms power recommendation engines
that provide personalized suggestions in areas such as e-commerce, content streaming
platforms, and social media.
7) Natural Language Processing (NLP): Machine learning enables the analysis and
understanding of human language. NLP is used for tasks like sentiment analysis,
language translation, chatbots, and voice assistants.

Abuses of Machine learning:

Abuses and Challenges of Machine Learning:
Bias and Discrimination: Machine learning algorithms can perpetuate biases present in the
training data, leading to unfair and discriminatory outcomes. Care must be taken to ensure
fairness and equity in algorithmic decision-making.
Privacy Concerns: Machine learning often requires large amounts of personal data, raising
concerns about data privacy and security. Safeguards must be in place to protect sensitive
information.
Lack of Transparency: Some machine learning algorithms, such as deep neural networks, can
be complex and difficult to interpret. The lack of transparency raises challenges in
understanding and explaining the decisions made by the algorithms.
Overreliance and Automation Bias: Blindly trusting machine learning algorithms without
human oversight can lead to errors or unintended consequences. Human intervention and
critical thinking are essential to prevent overreliance and automation bias.
Data Quality and Bias: Machine learning models heavily rely on high-quality and unbiased
training data. Biased or inaccurate data can lead to unreliable and unfair predictions.
Adversarial Attacks: Machine learning models can be vulnerable to adversarial attacks, where
malicious actors intentionally manipulate input data to deceive the model or cause it to make
incorrect predictions.
Job Displacement: Automation enabled by machine learning may lead to job displacement and
require reskilling and upskilling of the workforce to adapt to changing job roles.

How do machines learn:

Machines learn through the process of training on data and extracting patterns or relationships
from that data. The specific approach to machine learning depends on the algorithm or
technique used, but the general process involves the following steps:
Data Collection: The first step is to collect or obtain a dataset that represents the problem
domain. The dataset consists of input features (also called independent variables) and
corresponding target values (also called labels or dependent variables) for supervised learning,
or just input features for unsupervised learning.
Data Preprocessing: The collected data may require preprocessing to clean and transform it
into a suitable format for the learning algorithm. This step may involve handling missing data,
normalizing or scaling features, encoding categorical variables, or performing other data
transformations.
Training Data Split: The dataset is typically divided into two subsets: a training set and a test
set. The training set is used to train the machine learning model, while the test set is used to
evaluate the model's performance on unseen data.
Model Training: The training process involves feeding the training data into the machine
learning algorithm. The algorithm learns from the input features and the corresponding target
values (in the case of supervised learning) to find patterns or relationships that can be used to
make predictions or classifications.
Model Evaluation: After training, the model's performance is assessed using evaluation metrics
specific to the problem at hand. For example, in classification tasks, metrics like accuracy,
precision, recall, and F1 score may be used. In regression tasks, metrics like mean squared error
(MSE) or mean absolute error (MAE) are common.
Model Tuning: If the model's performance is not satisfactory, model tuning or hyperparameter
optimization may be performed. Hyperparameters are settings that influence the learning
process but are not learned directly from the data. By adjusting these hyperparameters, the
model's performance can be improved. Prediction or Inference: Once the model is trained and
evaluated, it can be used to make predictions or perform inferences on new, unseen data. The
trained model takes the input features of the new data and produces predictions or
classifications based on the learned patterns or relationships.
Model Deployment and Monitoring: The trained model is deployed into a production
environment, where it can be used to make real-time predictions or assist in decision-making.
Ongoing monitoring of the model's performance is important to ensure its accuracy and
reliability over time.

General Ml architecture:
The general architecture of a machine learning system typically involves the following
components:

 Data Collection and Storage:

o Data is collected from various sources such as databases, files, APIs, or web
scraping.
o The collected data is stored in a suitable storage system such as a database or
distributed file system.
 Data Preprocessing:
o Raw data is processed and cleaned to handle missing values, outliers, or
inconsistencies.
o Data normalization, scaling, or feature engineering techniques may be applied.
o The preprocessed data is transformed into a suitable format for machine
learning algorithms.
 Feature Extraction and Selection:
o Relevant features are extracted from the preprocessed data.
o Feature selection techniques may be applied to choose the most informative
features.
o Dimensionality reduction methods like Principal Component Analysis (PCA) may
be used.
 Model Selection and Training:
o An appropriate machine learning model is selected based on the problem and
data characteristics.
 o The selected model is trained using the preprocessed data. Training involves
optimizing model parameters using a suitable learning algorithm.
 Model Evaluation and Validation:
o The trained model is evaluated on a separate testing dataset.
o Evaluation metrics such as accuracy, precision, recall, or mean squared error are
calculated.
o Cross-validation techniques may be used for more robust evaluation.
 Model Deployment and Monitoring:
o The trained and evaluated model is deployed in a production environment.
o The model is integrated into an application or system to provide predictions or
recommendations.
o Continuous monitoring of the model's performance and metrics is important.
o Retraining the model periodically with new data may be necessary to maintain
accuracy.