BABAJI VIDHYASHRAM

SHOLINGANALLUR, CHENNAI – 600119

CLASS X AI NOTES
Chapter – 3 Modelling and Artificial Neural Network
Modelling Definition:
AI Modelling refers to developing algorithms, also called models which can be trained to get
intelligent outputs. That is writing codes to make a machine artificially intelligent.
Types of AI Models:

Rule Based Approach

Rule Based Approach refers to the AI modelling where the relationship or patterns in data are
defined by the developer. The machine follows the rules or instructions mentioned by the
developer, and performs its task accordingly.
Rule-based Chatbots are commonly used on websites to answer frequently asked questions
(FAQs) or provide basic customer support.

Disadvantages:
A drawback/feature for this approach is that the learning is static. The machine once trained,
does not take into consideration any changes made in the original training dataset. That is, if
you try testing the machine on a dataset which is different from the rules and data you fed it
at the training stage, the machine will fail and will not learn from its mistake. Once trained,
the model cannot improvise itself on the basis of feedbacks. Thus, machine learning gets
introduced as an extension to this as in that case, the machine adapts to change in data and
rules and follows the updated path only, while a rule-based model does what it has been
taught once.
Learning Based Approach
A learning-based approach is a method where a computer learns how to do something by
looking at examples or getting feedback, similar to how we learn from experience. Instead of
being explicitly programmed for a task, the computer learns to perform it by analyzing data
and finding patterns or rules on its own.

Categories of Machine learning based models

Learning-based approaches are indeed a broad category that encompass both machine
learning and deep learning. Machine learning can further be divided into three parts:

Supervised Learning
In a supervised learning model, the dataset which is fed to the machine is labelled. In other
words, we can say that the dataset is known to the person who is training the machine only
then he/she is able to label the data. A label is some information which can be used as a tag
for data. For example, students get grades according to the marks they secure in
examinations. These grades are labels which categorize the students according to their marks.
Supervised Learning –
Example
▪ Let’s consider the example of currency coins.
▪ Problem Statement: Build a model to predict the coin based on its weight.
▪ Assume that we have different currency coins(dataset) having different weights.
▪ 1 Euro weighs 5 grams
▪ 1 Dirham weighs 7grams
▪ 1 Dollar weighs 3grams
▪ 1 Rupee weighs 4 grams and so on
▪ Feature – Weights
▪ Label – Currency
Unsupervised Learning
An unsupervised learning model works on unlabelled dataset. This means that the data which
is fed to the machine is random and there is a possibility that the person who is training the
model does not have any information regarding it. The unsupervised learning models are used
to identify relationships, patterns and trends out of the data which is fed into it. It helps the
user in understanding what the data is about and what are the major features identified by the
machine in it.
For example, you have a random data of 1000 dog images and you wish to understand some
pattern out of it, you would feed this data into the unsupervised learning model and would
train the machine on it. After training, the machine would come up with patterns which it
was able to identify out of it. The Machine might come up with patterns which are already
known to the user like colour or it might even come up with something very unusual like the
size of the dogs.
Unsupervised Learning - Example
Let’s consider the example of a supermarket Assume that we have a customer database with
records of their products bought over a period. Now you being the marketing manager
decides to send a grocery offer message to those customers who buys grocery regularly.
• Note that there was no customer labelled as grocery shoppers and non-grocery
shopper.
• Model could discover patterns on its own and could come up with these two
clusters/groups.

Supervised Vs. Unsupervised Learning

Reinforcement Learning
This learning approach enables the computer to make a series of decisions that maximize a
reward metric for the task without human intervention and without being explicitly
programmed to achieve the task.
Reinforcement Learning – Example
▪ Reinforcement learning is a type of learning in which a machine learns to perform a
task through a repeated trial-and-error method.
▪ Let’s say you provide an image of an apple to the machine and ask the machine to predict it-
▪ The machine first predicts it as ‘cherry’ and you give negative feedback that it’s incorrect.
▪ Now, the machine learns that it’s not a cherry. Then again, you ask the machine to predict
the fruit by giving an image of an apple as input;
▪ Now, it knows it is not a cherry.
▪ It predicts it as an apple and you give positive feedback that it’s correct.
▪ So, now the machine learns that this is an apple.
Examples of Reinforcement Learning

Sub-categories of Supervised Learning Model

There are two types of Supervised Learning models: Classification model and Regression
model.
Classification Model
Here the data is classified according to the labels. For example, in the grading system,
students are classified on the basis of the grades they obtain with respect to their marks in the
examination. This model works on discrete dataset which means the data need not be
continuous.
Example of the Classification Model:
• Classifying emails as spam or not
Regression Model:
Such models work on continuous data. For example, if you wish to predict your next salary,
then you would put in the data of your previous salary, any increments, etc., and would train
the model. Here, the data which has been fed to the machine is continuous.
▪ Regression algorithms predict a continuous value based on the input variables.
▪ Continuous values as Temperature, Price, Income, Age, etc.

Examples of the Regression Model:

• Predicting Temperature
• Predicting the price of the house
• Used car price prediction
Sub-categories of Unsupervised Learning Model
Unsupervised learning models can be further divided into two categories: Clustering model
and Association model.

What is Clustering?
Clustering is a process of dividing the data points into different groups or clusters based on
their similarity between them.
Difference between Clustering and Classification
▪ Classification uses predefined classes in which objects are assigned.
▪ Clustering finds similarities between objects and places them in the same cluster and it
differentiates them from objects in other clusters.
Association
Association Rule is an unsupervised learning method that is used to find interesting
relationships between variables from the database.
Sub-Categories of Deep Learning
Deep Learning enables software to train itself to perform tasks with vast amounts of data. In
deep learning, the machine is trained with huge amounts of data which helps it to train itself
around the data. Such machines are intelligent enough to develop algorithms for themselves.
There are two types of Deep Learning models: Artificial Neural Networks (ANN) and
Convolution Neural Network (CNN).

Artificial Neural networks(ANN) - Artificial Neural networks are modelled on the human
brain and nervous system. They are able to automatically extract features without input from
the programmer. Every neural network node is essentially a machine learning algorithm. It is
useful when solving problems for which the data set is very large.
Convolutional Neural Network (CNN) - Convolutional Neural Network is a Deep Learning
algorithm which can take in an input image, assign importance (learnable weights and biases)
to various aspects/objects in the image and be able to differentiate one from the o.
What is Neural Network?
Neural networks are loosely modelled after how neurons in the human brain behave. The key
advantage of neural networks is that they are able to extract data features automatically
without needing the input of the programmer. A neural network is essentially a system of
organizing machine learning algorithms to perform certain tasks. It is a fast and efficient way
to solve problems for which the dataset is very large, such as in images.
Input Layer:
• The input layer acts as the entry point for data into the neural network.
• Each node in this layer corresponds to a feature in the input data.
• For example, if the input is an image, each node might represent a pixel's intensity
value.
Hidden Layers:
• These layers are located between the input and output layers.
• They perform computations and learn complex patterns from the input data.
• A neural network can have one or more hidden layers.
• Each layer receives input from the previous layer, processes it, and passes the result to
the next layer.
Output Layer:
• The output layer produces the final prediction or results of the neural network.
• The number of nodes in the output layer depends on the task. For example, a binary
classification problem might have one output node, while a multi-class classification
problem could have multiple nodes.
• For instance, in image classification, the output layer might have multiple nodes, each
representing a different class.

___________________________________________________________________________
Important Points to Remember
1. What is Modelling in AI?
Modelling is the process of creating a digital or virtual representation of a real-world problem
or system using data and logic. It helps understand, predict, or solve real-world problems
using machines.
Example: A weather forecasting app uses historical weather data to predict rain or sunshine.
2. Why is Modelling Important in AI?
- AI learns from data to make decisions.
- Modelling helps train AI systems to behave intelligently.
- Automates tasks like image recognition, chatbots, etc.
3. Types of Modelling
A. Rule-Based Modelling
Uses fixed if-else rules for simple decision-making.
Example:
if temperature > 30: turn_on_fan()
Advantages:
- Easy to understand
- No need for data
Limitations:
- Not suitable for complex/changing problems
- Cannot learn from new data
B. Data-Based Modelling
Uses machine learning algorithms to build models from data.
Example: Predicting loan default using historical loan data.
Advantages:
- Learns from data
- More accurate
Limitations:
- Requires lots of data
- Can be complex to interpret
4. What is Artificial Neural Network (ANN)?
An ANN is a machine learning model inspired by the human brain. It recognizes patterns,
classifies data, and makes predictions using interconnected processing units called neurons.
5. Structure of ANN
ANN has three layers:
- Input Layer: Accepts raw data
- Hidden Layer(s): Processes data
- Output Layer: Produces result
6. How Does ANN Work?
1. Input is provided
2. Data flows through hidden layers using weights
3. Activation functions decide outputs
4. Output is generated
5. Model learns using backpropagation if errors occur
7. Human Brain vs Artificial Neural Network
Human Brain:
- Uses biological neurons
- Learns through experience
- Processes emotions

ANN:
- Uses artificial neurons
- Learns from data
- Processes only numbers
8. Applications of ANN
- Healthcare: Disease detection
- Banking: Fraud detection
- Email: Spam filtering
- Retail: Product recommendations
- Transportation: Self-driving cars
- Social Media: Facial recognition
9. Key Terms to Remember
- Neuron: Basic processing unit
- Weights: Importance of input
- Activation Function: Decides neuron output
- Training: Learning from data
- Backpropagation: Adjusting weights
- Model: Final trained ANN
10. Example: Email Classification
Input: Email content
Hidden Layer: Processes features (e.g. keywords)
Output: Classifies as Spam or Not Spam
11. Steps to Build ANN Model
1. Collect Data
2. Prepare Data
3. Train the Model
4. Test the Model
5. Evaluate Accuracy
12. CBSE Focus Points
- Know types of modelling
- Understand ANN structure
- Recognize real-life applications
- Learn how ANN is trained using data.
CRITICAL THINKING QUESTIONS

Scenario 1: Health Monitoring App

You are building an AI model for a health app that predicts a person's risk of diabetes.
Q1: The model is showing high accuracy during training but poor performance when tested
on new users. What might be the problem?
Answer: The model may be overfitting—it learned the training data too well but failed to
generalize to new data. This means it’s not adaptable to real-world cases.
Q2: The data you collected is mostly from urban users aged 25–40. How might this affect the
model's predictions for rural users or older adults?
Answer: The model may become biased and give inaccurate predictions for
underrepresented groups (e.g., rural users or older people) because it hasn’t learned patterns
from diverse data.
Q3: A parent asks if the model is always right. How would you explain the concept of
"probability" and "prediction" to them?
Answer: AI models don’t give guaranteed answers; they give probabilities based on patterns
in the data. A high risk does not mean certainty—it’s just a prediction based on known trends.

Scenario 2: Movie Recommendation System

Your school is designing an AI system to recommend movies to students.
Q1: One student complains that the app only shows similar kinds of movies. What modelling
issue could be causing this?
Answer: This may be due to a narrow training dataset or a lack of diversity in the
algorithm, causing the model to recommend only what it already "knows" you like—a
problem called filter bubble.
Q2: If the data used to train the system includes only English-language movies, what impact
could that have?
Answer: The model may fail to recommend regional or foreign language movies, making it
less useful or inclusive for students with diverse preferences.
Q3: How could you modify the model to better include diverse student preferences?
Answer: Include more varied training data from different genres, languages, and cultures.
Use user feedback loops to help the model learn new interests over time.

Scenario 3: Handwriting Recognition

You're helping develop an ANN to read handwritten test answers.
Q1: If the ANN misreads similar letters (like 'a' and 'o'), what part of the ANN or training
process might need improvement?
Answer: The training data may need more examples of these letters, especially in different
handwriting styles. The feature extraction process could also be enhanced.
Q2: How could increasing the number of hidden layers affect the performance of the model?
Answer: More hidden layers allow the ANN to learn more complex patterns, potentially
improving accuracy. However, it may also increase training time and risk overfitting.
Q3: A student writes in cursive and the ANN can’t recognize it. How can the training data be
improved to fix this?
Answer: Add cursive handwriting samples to the training data to help the ANN learn and
adapt to different writing styles.

Scenario 4: Voice-Controlled Assistant

You’re developing a voice assistant for classroom use.
Q1: The assistant struggles to understand accents and background noise. How could you use
modelling or ANN to improve it?
Answer: Use a diverse voice dataset including various accents and noisy environments.
Train the ANN with noise-robust features to handle real-world audio better.
Q2: If the ANN gives a wrong response, how would you decide whether it’s a data issue,
model issue, or training issue?
Answer: Analyze logs and performance:
• If data is unbalanced or low quality → data issue
• If model architecture is too simple or complex → model issue
• If performance worsens after more training → training issue (like overfitting)
Q3: Why might using only a small set of training voices limit the assistant's real-world
performance?
Answer: The ANN will not generalize well to unseen voices or speech patterns, leading to
poor recognition in real usage. It needs a large, diverse voice dataset.