PART-B – SUBJECT SPECIFIC SKILLS

Unit 1: AI Reflection, Project Cycle and Ethics

LEARNING OUTCOMES:
Learners will be able to
• Identify and appreciate Artificial Intelligence and describe its applications in
daily life.
• Relate, apply and reflect on the Human-Machine Interactions to identify and
interact with the
o three domains of AI: Data, Computer Vision and Natural Language
Processing and Undergo
o assessment for analyzing their progress towards acquired AI-
Readiness skills.
• Get familiar and motivated towards Artificial Intelligence and Identify the
AI Project Cycle framework.
• Learn problem scoping and ways to set goals for an AI project and
understand the iterative nature of problem scoping in the AI project cycle.
• Identify stakeholders involved in the problem scoped.
• Understand the iterative nature of problem scoping for in the AI project
cycle.
• Brainstorm on the ethical issues involved around the problem selected.
• Foresee the kind of data required and the kind of analysis to be done.
• To understand the purpose of Data Visualization
• Use various types of graphs to visualize acquired data.
• Identify data requirements and find reliable sources to obtain relevant data.
• Understand various evaluation techniques.
Introduction to AI
Artificial intelligence (AI) refers to the capability of computational systems to
perform tasks typically associated with human intelligence, such as learning,
reasoning, problem-solving, perception, and decision-making. The term "artificial
intelligence" was coined by John McCarthy in 1956.
Applications of AI
Advanced web search engines (e.g., Google Search); recommendation systems (used
by YouTube, Amazon, and Netflix), virtual assistants (e.g., Google
Assistant, Siri, Alexa), autonomousvehicles (e.g., Waymo); generative and creative

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 61

tools(e.g., ChatGPT , Deepseek, Grok,AI art) ,Health Care (AI driven medical
imaging ),Social media(personalized recommendations, content moderations, content
creation),Banking(Fraud and Risk detection) ,Agriculture( AI powered tools to
optimize resources use and crop management),Space explorer( safe landing of
Chandrayan-III on lunar south pole).
How to make machine Intelligent?
When a machine can learn, reason and adapt.
DATA+ALGORITHM=AI MACHINE

Domains of AI
specialized areas within artificial intelligence that deal with specific problems,
techniques, and applications.

Domains of AI

Statistical Data
CV NLP
Statistical Data:
Statistical Data refers to statistical techniques to analyze, interpret and draw insights
from numerical/tabular data
• It is the backbone of AI here the data is called bigdata (voluminous data).
• Data is derived from various sources private as well as public.
• The data is processed by analyzing patterns and trends to speculate future
predictions.

Computer Vision (CV): Computer Vision, is an AI domain works with videos

and images enabling machines to interpret and understand visual information.
• It uses machine learning and neural networks to enable computers and systems
to understand and interpret the visual world. It replicates the complexity of
human vision.
• Computer vision enables machines to "see" and interpret images, much like
humans.

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 62

 Natural language processing (NLP)
AI domain focused on textual data enabling machines to comprehend, generate,
and manipulate human language. i.e. computers to understand human language,
whether written, spoken, or even scribbled.
Some AI Applications
• Face Lock in Smartphones
Smartphones nowadays come with the feature of face locks in which the
smartphone’s owner can set up his/her face as an unlocking mechanism for it.
• Smart assistants
Smart assistants like Apple’s Siri and Amazon’s Alexa recognize patterns in
speech, then infer meaning and provide a useful response.
• Fraud and Risk Detection
Finance companies were fed with bad debts and losses every year. However, they
had a lot of data which used to get collected during the initial paperwork while
sanctioning loans.
• Medical Imaging: For the last decades, computer supported medical
imaging
application that has been a trustworthy help for physicians. It doesn’t only create
and analyze images, but also becomes an assistant and helps doctors with their
interpretation.
AI Project Cycle
AI Project Cycle is a structured approach to developing AI solutions, involving
six key stages:
● Problem Scoping
● Data Acquisition
● Data Exploration
● Modeling
● Evaluation
● Deployment
Use of AI Project Cycle
AI project cycle is the cyclical process followed to complete an AI project.
AI project cycle helps us:
• to understand the process

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 63

 • to create better AI projects easily
• to create AI projects faster

1. Problem Scoping: This initial stage involves clearly defining the problem
you're trying to solve, setting project goals, and identifying the key stakeholders,
identifying existing measures, identifying ethical issue which will influence the
project.
4W frameworks:
Who: Stakeholders (who are facing the problems and wants a solutions)
What: What is the actual problem. To collect evidence to prove that the problem
actually exists.
Where: Context/situation/location of the problem.
Why: How the solution will be benefited to the society

Problem Statement Template

The Problem Statement Template helps us to summarize all the key points into
one single Template so that in future, whenever there is a need to look back at
the basis of the problem, we can take a look at the Problem Statement Template
and understand the key elements of it.

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 64

 2. Data Acquisition: (Acquiring Data from reliable sources)
In this stage, you collect the relevant data that will be used to train and evaluate
your AI model.
Types of Data:
Training Data: collecting and preparing the data needed to train the AI
model.
Testing Data: Testing data is a separate dataset used to evaluate the performance
of attained model after it has been developed.

Various sources for data acquisition:

o Data can be acquired from a wide range of sources, including physical
sensors, online databases, social media platforms, and even by
purchasing datasets from external vendors.
o These sources can be categorized into physical sensors, online
databases, and data warehouses.
o Physical Sensors and Transducers: These directly measure physical
phenomena like temperature, pressure, and movement.
o Online Databases: This includes structured data stored in relational
databases, APIs (Application Program Interface), and flat files.
o Data Warehouses: These are centralized repositories that store data
from various operational systems, databases, and external sources like
partner systems, IoT devices, weather apps, and social media.
Sharing/Exchanging Data: Data can be shared or exchanged with
other organizations or entities.
Public Datasets: Free datasets from government or research

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 65

 repositories can be used.
Internal Databases: Organizations can access data from their own
proprietary databases or internal sources.
IoT Devices: Data can be acquired from IoT devices connected to
operational systems.

3. Data Exploration: This involves analyzing and visualizing the collected data
to gain insights, understand its characteristics, and identify potential issues.

Data visualization techniques which you learnt from :-

https://datavizcatalogue.com
Steps in data exploration:
Data Collection: Gathering data from various sources.
Data Cleaning: Addressing missing values, duplicates, and inconsistencies.
Data Transformation: Converting data into a suitable format for analysis.

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 66

 Data Visualization: Using graphs, charts, and other visuals to explore patterns
and relationships.
Statistical Analysis: Applying statistical techniques to analyze the data.
Feature Selection: Identifying relevant features for modeling.
Documentation: Documenting the data exploration process and findings

4. Modeling:
In this module, progress from data exploration to AI modeling, learning
about key distinctions between Artificial Intelligence (AI), Machine Learning (ML),
and Deep Learning (DL). The module introduces two approaches to AI modeling:
Rule-Based and Learning-Based

This is the core stage where you develop and train the AI model to learn patterns
from the data and make predictions or decisions using algorithms and mathematical
frameworks to solve a defined problem

AI, ML & DL
AI (Artificial Intelligence)-refers to any technique that enables computers to
mimic human intelligence. An artificially intelligent machine works on algorithms
and data fed to it and gives the desired output.
ML (Machine Learning) - Enables machines to improve at tasks with experience.
The machine here learns from the new data fed to it while testing and uses it for the
next iteration. It also takes into account the times when it went wrong and
considers the exceptions too.
DL (Deep Learning)- Enables software to train itself to perform tasks with vast
amounts of data. Since the system has got huge set of data, it is able to train itself
with the help of multiple machine learning algorithms working altogether to
perform a specific task.
“Artificial Intelligence is the
umbrella term which holds both Deep
Learning as well as Machine Learning.
Deep Learning, on the other hand, is the
very specific learning approach which
is a subset of Machine Learning as it
comprises of multiple Machine
Learning algorithms.”

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 67

Rule Based Approach
Refers to the Al modeling where the rules are defined by the developer. The machine
follows the rules or instructions mentioned by the developer and performs its task
accordingly.
For example, we have a dataset which tells us about the conditions on the basis of
which we can decide if child can go out to play golf or not.
The parameters are: Outlook, Temperature, Humidity and Wind

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 68

Learning Based Approach
Refers to the Al modelling where the machine learns by itself. Under the Learning
Based approach, the Al model gets trained on the data fed to it and then is able to
design a model which is adaptive to the change in data.
1. Supervised Learning:
This type of learning uses labeled data, where each input comes with a
corresponding correct output, to train a model
Examples:
Image classification: Training a model to identify different objects in
images, like cats, dogs, or cars.
Spam detection: Training a model to distinguish between spam and
legitimate emails.
Medical diagnosis: Training a model to identify different types of tumors in
medical images.
Predicting stock prices: Training a model to predict the future price of a
stock based on historical data.
2. Unsupervised Learning:
This type of learning uses unlabeled data to discover patterns and relationships
within the data.
Examples:
Customer segmentation: Grouping customers with similar purchasing
behaviors for targeted marketing.
Anomaly detection: Identifying unusual or suspicious patterns in data, such
as fraudulent transactions.
Feature extraction: Identifying the most important features in a dataset for
analysis.

Benefits of Learning-Based Approach

Flexibility: Can handle a wide range of problems and adapt to new

situations.
Adaptability: Can improve its performance over time with more data.
Pattern Recognition: Can identify complex patterns and relationships in
Data.

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 69

 5. Evaluation:
This final stage assesses the performance of the trained model by comparing its
predictions to actual results and evaluating its reliability. Evaluation is the
process of understanding the reliability of any AI model, based on outputs by
feeding test dataset into the model and comparing with actual answers. There can
be different Evaluation techniques, depending of the type and purpose of the
model.

Purpose of Evaluation

Understanding Model Performance:

Evaluation helps determine how well the AI model can solve the problem it was
designed for.
Identifying Strengths and Weaknesses:
By analyzing the model's performance, you can identify areas where it excels
and areas where it needs improvement.
Guiding Model Refinement:
The evaluation results can inform further adjustments to the model, such as
retraining with different algorithms or adjusting parameters.

Key Evaluation Metrics:

Accuracy:
Measures the overall correctness of the model's predictions, encompassing both
true positives and true negatives.
Precision:
Measures how well the model avoids false positives (incorrectly identifying
something as positive when it's not).
Recall (Sensitivity):
Measures how well the model correctly identifies all positive cases, avoiding
false negatives (missing positive cases).

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 70

 F1-score:
Combines precision and recall, providing a balanced view of the model's
performance, especially useful when dealing with imbalanced datasets.

Model Evaluation Terminologies

Imagine that you have come up with an AI based prediction model which has
been deployed in a forest which is prone to forest fires. Now, the objective of
the model is to predict whether a forest fire has broken out in the forest or not

Here we have two conditions (Prediction and Reality)

Prediction (output which is given by the machine) and Reality (real scenario
in the forest when the prediction has been made)

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 71

 • True Positives (Correct Positive Predictions)
• True Negatives (Correct Negative Predictions)
• False Positives (Incorrectly Predicted Positives)
• False Negatives (Incorrectly Predicted Negatives)

6.Deployment
Deployment as the final stage in the AI project cycle where the AI model or
solution is implemented in a real-world scenario.
Key steps involved in the deployment process:
a. Testing and validation of the AI model
b. Integration of the model with existing systems
c. Monitoring and maintenance of the deployed model.
Some examples of successful AI projects that have been deployed in various
industries, such asself-driving cars, medical diagnosis systems, and chatbots.

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 72

 AI ETHICS
Refers to the principles that guide the development and use of AI systems to
ensure they are beneficial, safe, and responsible. These principles consider
fairness, transparency, accountability, privacy, security, and potential societal
impacts of AI.
Difference between ethics and morals:
• Ethics generally refers to societal rules and standards, while morals are
more about personal beliefs and values
• Ethics are external, while morals are internal (Ethics are often enforced by
legal or organizational sanctions, while morals are more personal).
Examples:
Moral: A business owner choosing to donate a portion of their profits to
charity because they believe in giving back to the community.
Ethics: A business owner to donate some percentage of his profit under
CSR (Company Social Responsibility)
Ethics can include avoiding bias, ensuring privacy of users and their data, Inclusion
and environmental risks.

Ethics vs Morals

Morals Ethics
• The beliefs dictated by our • The guiding principles to decide
society. what is good or bad.
• Morals are not fixed and can • These are values that a person
be different for different themselves chooses for their life
societies Examples:
Examples: • Is it good to speak the truth in all
• Always speak the truth situations?
• Always be loyal • Is it good to be loyal under all
• Always be generous circumstances?
• Is it necessary to always be
generous?

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 73

Principles in AI Ethics
To make AI better, we need to identify the factors
responsible for it.
The following principles in AI Ethics affect the quality
of AI solutions
Here are a few things that you should take care of :
Human Rights
• Does your AI take away Freedom?
• Does your AI discriminate against People?
• Does your AI deprive people of jobs?
• What are some other human rights which need to be protected when it comes
to AI?
Bias
• Does your data equally represent all the sections of the included
populations?
• Will your AI learn to discriminate against certain groups of people?
• Does your AI exclude some people?
• What are some other biases that might appear in an AI?
Privacy
• Does your AI collect personal data from people?
• What does it do with the data?
• Does your AI let people know about the data that it is collecting for its use?
• Will your AI ensure a person’s safety? Or will it compromise it?
• What are some other ways in which AI can breach someone’s privacy?
Inclusion
• Does your AI collect personal data from people?
• What does it do with the data?
• Does your AI let people know about the data that it is collecting for its use?
Will your AI ensure a person’s safety? Or will it compromise it?
• What are some other ways in which AI can breach someone’s privacy?

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 74

Ethical scenarios faced while building AI solutions:
Scenario 1: If a driverless car finds someone crossing the road, takes a turn to
avoid hitting the person and instead smashed the car in a wall / tree nearby
causing serious injury to the passengers, is it a fault in the development on
moral ground.
Scenario 2: If any Automated or Driverless Car hits someone standing in the
middle of a road, who needs to be held responsible; whom should be penalized

AI bias
The biased mentality of the developers may result in the presence of bias in the
AI System. Bias might not be negative always but it may reflect the social
norms and common facts.
AI bias can develop due to the data used to train the AI
Examples of AI bias:
All the virtual assistants have a female voice.
• If searching is done for Salons on Google, the resultant list mostly contains
female salons.
• Searching for images of Nurse gives the pictures of female nurses mostly.
Some recent examples of AI bias?
• Racial AI Bias in the U.S. Healthcare
They discovered that a widely used healthcare algorithm, affecting over 200 million
patients in U.S. hospitals, significantly favored white patients over Black patients
when predicting who needed extra medical care.
AI Access
• AI Access refers to the ability to securely and appropriately use artificial
intelligence tools and applications.
• Since AI Systems are emerging technology and expensive, all cannot get
access to AI and get benefitted. Ex-people belong to EWS may not be
benefitted from AI
• AI may create unemployment if not implemented carefully and with
compassion.

KVS ZIET MYSORE AI SSM(417) 2025-26 Page 75