WEST BENGAL COUNCIL OF HIGHER SECONDARY EDUCATION

SYLLABUS FOR CLASS XI AND XII

SUBJECT : ARTIFICIAL INTELLIGENCE ( ARTI )

Course Objectives :

The objectives of this course are:

● To impart knowledge about basic computer fundamentals and programming environments

required for implementing Artificial Intelligence (AI) systems.
● To enable the students to understand the history of AI and the basic principles of modern AI.
● To enable students to learn the informed and uninformed search methods, and a simple
evolutionary algorithm for solving problems.
● To enable students to understand the fundamentals of knowledge representation, building of
simple knowledge-based systems, and to apply knowledge representation and reasoning
● To enable the students to understand important machine learning(ML) algorithms used for
improving various components of an AI agent
● To enable the students to understand the uses of AI and ML in various applications.
● To enable the student to understand ethics in AI
● To gain practical experience in developing various AI and ML models

Course outcomes:

Upon successful completion of this course, the student shall be able to:

● Demonstrate an understanding of the history of AI and its foundations.

● Apply basic principles of AI in problem-solving that require perception, knowledge representation ,
inference, and learning.
● Demonstrate awareness and a fundamental understanding of various applications of AI and
Machine Learning techniques in real-world problem solving.
● Demonstrate proficiency in developing various real-world AI and ML applications using the latest
programming languages and software tools.
● Demonstrate an ability to share in discussions of AI and ML, its current scope and limitations, and
its impact on society.
 CLASS - XI
SEMESTER – I
SUBJECT: ARTIFICIAL INTELLIGENCE ( ARTI )
FULL MARKS: 35 CONTACT HOURS: 60 Hours
COURSE CODE: THEORY

SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
History of computer, Basic Computer hardware, input and output

devices, Basic computer architecture, input output devices, memory

and CPU, networking of machines (overview of LAN, MAN, WAN,

Internet, Wifi etc), types of computer (workstation, desktop,

1a Smartphone, embedded system, etc.), Overview of Software (system 8 5

software and application software with examples (mention names

only), Definition of Operating System and functions (mention names

of some popular operating systems like Windows, Linux, Android,

etc).

Bit, Byte and Word, Number System (Base, Binary, Decimal, Octal,
Unit -1 Hexadecimal), Conversion of number systems, Boolean logic
Computer 6 5
1b
Fundamental (Boolean Gates ), Boolean operators (OR, AND and NOT), ASCII code,
(15)
Concept of Algorithm and Flowchart.

Basics of Computer Programming (three levels: high level language,

assembly language, machine language, definition and block

diagrams), Overview of Compiler and Interpreter (definition and

mention name of major compiled (e.g., C, C++) and interpreted

1c languages (e.g., Python), Overview of procedural and object 10 5

oriented programming (key features and just the basic differences,

mention names of some popular procedural (e.g., BASIC,

FORTRAN, C) and object oriented programming languages (e.g., C++,

Java, Python).
 SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
Basics of Python programming (with a simple 'hello world' program,

process of writing a program, running it, and print statement),

Concept of class and object, Data-types (integer, float, string), Notion

of a variable, Operators (assignment, logical, arithmetic etc.),

2a accepting input from console, conditional statements (If else and 12 5

Nested If else ), Collections (List, Tuple, Sets and Dictionary), Loops

(For Loop, While Loop & Nested Loops), Iterator, String and

fundamental string operations (compare, concatenation, sub-string

etc.), Function, Recursion.

Unit -2
Overview of linear and non-linear data structure (definition,
Introduction
to Python schematic view and difference), array (1D, 2D and its relation with
Programm-
matrix, basic operations: access elements using index, insert, delete,
ing (15)
search), stack (concept of LIFO, basic operations: Push, Pop, peek,

2b size), queue (concept of FIFO, basic operations: Enqueue, Dequeue, 12 6

peek, size), use of List methods in python for basic operations on

array, stack and queue, overview of NumPy library and basic array

operations (arrange(), shape(), ndim(), dtype() etc.), binary tree

(definition and schematic view only) .

Linear search and binary search algorithm, sorting algorithm

2c 4 4
( bubble sort only)

Basic matrix operations like matrix addition, subtraction,

multiplication, transpose of matrix, identity matrix. A brief

Unit -3 introduction to vectors, unit vector, normal vector, Euclidean space

Introduction Probability distribution, frequency, mean, median and mode, variance 8 5
3
to Linear
Algebra (5) and standard deviation, Gaussian distribution, Distance function,

Euclidean norm, distance between two points in 2D and 3D and

extension of idea to n dimensions

NB : Additional 10 hours for Remedial and/or Tutorial classes

 CLASS - XI
SEMESTER – II
SUBJECT: ARTIFICIAL INTELLIGENCE ( ARTI )
FULL MARKS: 35 CONTACT HOURS: 60 HOURS
COURSE CODE: THEORY

SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
History of AI: Alan Turing and cracking enigma, mark 1 machines,
1956-the birth of the term AI,
AI winter of 70’s, expert systems of 1980s, skipped journey of
present day AI. Distinction between terms AI, Pattern recognition
4a 6 3
and Machine Learning
Note: should be taught as a story more than flow of information
World war 2, Enigma and Alan Turing, the birth of modern
computers.

Search as optimization: how to search for the best answer to a

question? playing tic-tac-toe
● State Space Search, different states as different solutions of a
problem
● Mathematical equation for optimizing a result, example tic-tac-
toe, the states of the board and equation to calculate score of
the board with respect to a player
● Expanding possible states from a state and choosing the best
Unit -4 state
4b 16 10
Foundation of Uninformed search
AI & Search as a) Breadth first search
Optimization b) Depth first search
(18) Informed search
a) Heuristic search strategy with tic tac toe example
b) Greedy best-first search
c) A* search - basic idea only( without proof)
d) Hill climbing (only basic idea with a small example)
e) Simulated Annealing (No algorithm, Only basic idea)

Evolution and Darwin's theory, inspiration of evolutionary

algorithms, crossover and mutation,
Russian roulette for random selection, optimization using genetic
algorithm, one use of GA (to be chosen) practical: mention libraries
and problem.
4c ● Natural evolu on theory, survival of the fi est 10 5
● Expressing a solu on vector as gene, example of binary strings
● Crossover and muta on, its equivalent over binary strings
● Random selec on of genes from pool and random muta on
● Fitness func on
● Prac cal example by finding the root of a univariate equa on.
 SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
Logic in computer science, propositional logic, logic as expressions,
truth table, conjunction, disjunction, syllogism, tautology, De
Morgan's theorem. Use of logic to derive conclusions with practical
examples [NO LAB COMPONENT]
● Statements as logical proposi ons
● Atomic and compound proposi ons
5 ● Nega on, conjunc on and disjunc on as NOT, AND and OR
Knowledge ● Implica on and Bicondi onal statements
representation 5 ● Truth table as a way of proving propositions 16 10
and reasoning ● Commuta vity and Associativity and Distributive rules
(10) ● De Morgan's theorem
● Prac cal examples to infer meanings from statements
● Simple concept of Unifica on ( without details of MGU)
● Simple concept of clause (With Simple example)
● Basic concept of Inference (With Simple example)
● Example of Answer Extrac on system
● A brief introduc on to fuzzy logic (Only basic idea )

Handling Uncertain Knowledge

Uncertainty and Rational decision
6 Probabilistic Reasoning
Uncertainty Bayes Rule
6 8 5
Management Conditional probability
(5) Probabilistic inference using Bayes rule
a. General method(Simple cases)
b. Combining evidence

7 What is Chatbot? 4 2
Preliminary • Examples of different Chatbots
7
Concept of • The flowchart describing basic working principle of Chatbots.
Chatbots (2)

NB : Additional 10 hours for Remedial and/or Tutorial classes

 CLASS: XI
SUBJECT: ARTIFICIAL INTELLIGENCE ( ARTI )
COURSE CODE: PRACTICAL
FULL MARKS: 30 CONTACT HOURS: 60 HOURS

Sub Topic

SL CONTACT
TOPICS MARKS
NO HOURS

1. Computer Fundamentals [ No marks ]

● Visit to Computer Lab and familiarization with computers and

peripherals and different networking devices (e.g., modem, switch,
router).
1a 6 0
● Opening of the CPU box/cabinet and identification of different parts
(e.g., Motherboard, CPU/Processor, RAM, Hard Disk, power supply).

2. Introduction to Python Programming [15 Marks ]

● Introduction to installation and running of python codes with hello

world and simple accessing user inputs from console examples.
● Menu driven arithmetic calculator
● Simple logical and mathematical programs (e.g., printing patterns,
Conversion of binary to decimal and vice versa, Computing GCD of two
2a 12 3
numbers, Finding prime numbers, Generating Fibonacci sequence,
Computing factorial –iterative and recursive etc.)
● Finding max, min, avg, sum, length of a list
● Use of basic string methods like upper(), lower(), count(), find(), join(),
replace(), split() etc.

● Use of Python List methods for Stack and Queue implementation, for
examples, append() and pop()
● Use of NumPy array methods: arrange(), shape(), ndim(), size(), add(),
2b 6 7
subtract(), multiply(), divide(), mat() etc.
● Use of NumPy matrix multiplication methods: dot(), matmul(),
multiply() etc.
 SL CONTACT
TOPICS MARKS
NO HOURS

● Linear search and binary search in an array

2c 4 5
● Bubble sort in an array

3. Foundation for AI [ 3 Marks ]

● Generation of random numbers in python following a Gaussian

distribution and filling up random arrays
● Introduction to matplotlib to plot arrays as histograms
3a 8 3
● Computation of mean, median and mode
● Plotting Gaussian distribution with a given mean and standard
deviation

4. Search as Optimization (basic principles and example based understanding) [ 12 Marks ]

• Implementing 8-puzzle problem using DFS and BFS

• Use of class to denote state of a problem, example board state of tic
tac toe
4a
• Expansion of possible states from a given state with all possible moves 20 7
• Score function of each state and selection of highest score or least cost
at each level, i.e. making game tree

• Use of a class with an array and fitness score to define a solution of a

problem, example root finding of a linear equation with the solution
4b stored in binary. Basic genetic algorithm 4 5
• Binary crossover and mutation, selection using CDF of random
distribution, i.e Russian Roulette

NB : Additional 10 hours for Remedial and/or Tutorial classes

 CLASS - XII
SEMESTER – III
SUBJECT: ARTIFICIAL INTELLIGENCE ( ARTI )
FULL MARKS: 35 CONTACT HOURS: 60 Hours
COURSE CODE: THEORY

SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
Distance between distributions - Euclidean norm, Pearson correlation
coefficient, basic concepts of (not in details) chi square distance, Bayes
theorem and Bayesian probability
1
● Real n dimensional space (R^n) and Vector Algebra ,dot product of
Foundation
two vectors, vector projections.
of Statistics 1 10 5
● Product moment correlation coefficient (Pearson's coefficient) its use
for Machine
in determining relation between two sets of data
Learning (5)
● Chi square and , use in finding distance between two distributions
● Conditional probability and Bayes theorem , conditional
independence
• What is machine learning?
• Difference between traditional programming and Machine Learning
• Relation of Machine Learning with AI
• Applications of machine learning
• Why should machines have to learn? Why not design machines to
perform as desired in the first place?
• Types of Machine Learning (Supervised, Unsupervised, Semi-
supervised and reinforcement learning)
• Linear Regression with one variable
2a • Hypothesis representation, hypothesis space 16 10
2 • Learning requires bias
Introductio • Concept of training examples
n to • Concept of Loss function ,
Machine • Training methods: Iterative trial-and-error process that machine
Learning learning algorithms may use to train a model, Disadvantages of
(15) iterative training method,
Mean Squared Error(MSE), Gradient descent algorithm.
Effect of learning rate on reducing loss,
Importance of feature scaling (min-max normalization).

What is a feature or attribute? Some examples

Types of features(continuous, categorical)
Representation of training examples with multiple features
2b 8 5
Linear Regression with multiple attributes (Only formula for finding
weight vector without Derivation)
Feature cross and Polynomial Regression
 SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
• Difference between regression and classification.
• Examples of some real world classification problems,
• Linear classification and threshold classifier, Concept of
misclassification error, accuracy.
• Concept of input space and linear separator
• Drawback of threshold classifier
• Logistic regression model (without derivation)
3. 3a 18 10
• Use of logistic function in defining hypothesis function for logistic
Supervised regression model.
Learning • Probabilistic interpretation of output of the logistic regression model
(15) • Use of logistic regression model in binary classification task.
• Multi-class classification using One vs. all strategy.
• Instance based learning, K-nearest neighbor classifier, curse of
dimensionality

Measuring Classifier performance: confusion matrix, true positive, true

3b negative, false positive, false negative, error, accuracy, precision, recall, 8 5
F-measure, sensitivity and specificity, K-fold cross validation

NB : Additional 10 hours for Remedial and/or Tutorial classes

 CLASS - XII
SEMESTER – IV
SUBJECT: ARTIFICIAL INTELLIGENCE ( ARTI )
FULL MARKS: 35 CONTACT HOURS: 60 HOURS
COURSE CODE: THEORY

SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
What is supervised learning

Name some of the supervised learning algorithm

Probabilistic classifier:

Basics of Bayesian Learning, Conditional independence, Naive Bayes

classifier. Applications of Naive Bayes Classifier to sentiment

classification task, add-one smoothing.

Decision tree Learning:

Concept of entropy for measuring purity (impurity) of a collection of

4a training examples. 20 10

and information gain as a measure of the effectiveness of an attribute in

4 classifying the training data (just basics and equation) .

Unsupervised Inducing decision tree from the training data using the ID3 algorithm , an

Learning illustrative example showing how the ID3 algorithm works.

(15) Concept of overfitting, reduced error pruning,

Discretizing continuous-valued Attributes using information gain-based

method

(binary split only)

• What is unsupervised learning?

• Difference between supervised and unsupervised learning.

• What is clustering?

4b • Why is clustering and unsupervised learning technique? 12 5

• Some examples of real world application of clustering,

• Difference between clustering and classification.

• K-means clustering algorithm. Simple use cases

 SUB CONTACT
UNIT NO. TOPICS MARKS
UNIT HOURS
● Biological motivation for Artificial Neural Networks(ANN)
● A simple mathematical model of a neuron (McCulloch and
Pitts(1943))
● Concept of activation function: threshold function and Sigmoid
function,
● Perceptron as a linear classifier, perceptron training rule
● Representations of AND and OR functions of two inputs using
threshold perceptron.
● Equation of a linear separator in the input space, Representational
power of perceptrons
● Training unthresholded perceptron using Delta rule(with
derivation) , Need for hidden layers , XOR example,
5
● Why do we need non-linearity? Network structures: feed forward
Artificial Neural
5a networks and recurrent networks (basic concept only) 26 17
Network
● Training multiplayer feed-forward neural networks using Back
(17)
propagation algorithm (Concepts only and no derivation).
● Generalization, overfitting, and stopping criterion, overcoming the
overfitting problem using a set of validation data
● An Illustrative example of an ANN architecture for handwritten
digit recognition (Only input representation, output
representation and a block diagram of the network)
● Need for automatic feature learning, difference between the
conventional feed-forward neural networks and CNN, role of
convolution layer in CNN, An example of 2D convolution, function
of pooling layer
● A block diagram illustrating CNN applied to handwritten digit
recognition task

6 Brief discussion on important ethical issues in AI 2 3

Ethics in AI 6a
(3)

NB : Additional 10 hours for Remedial and/or Tutorial classes

 CLASS: XII
SUBJECT: ARTIFICIAL INTELLIGENCE ( ARTI )

COURSE CODE: PRACTICAL

FULL MARKS: 30 CONTACT HOURS: 60 HOURS

SL CONTACT
TOPICS MARKS
NO HOURS

1. Foundation of Statistics for Machine Learning [ 2 Marks ]

● Calculating Euclidean distance between two vectors using a python

program without using any library.

● Consider a table of data about n persons with two attributes-age and

1a 8 2
income and find Pearson correlation coefficient using a python program.

Do not use any ready -to-use library function that directly accepts the data

table and produces the output.

2. Introduction to Machine Learning [ 6 Marks ]

● Introduction to python libraries like scipy and statsmodel to various basic

codes

● Revisit matrix operations using scipy (basic matrix operations of addition,

2a subtraction, 12 3

multiplication, transpose)

● Using Scipy for advanced matrix operations - inverse

• Generation of random (x, y) pairs where y = f(x) + d (d varies from -r to +r ,

a random value ), f being a linear function

2b • Linear regression or line fitting of the data 6 3

• Optimizing the function using gradient descent

• Plotting the steps using matplotlib

SL CONTACT
TOPICS MARKS
NO HOURS

3. Supervised Learning [ 13 Marks ]

● Building linear regression-based threshold classifier and testing the model

on Diabetes Data set downloadable from UCI Machine Learning Repository
● Building Logistic regression model for binary classification of Diabetes
Data. Vary learning rate and verify the impact of learning rate on
classification performance.
● Introduction to the IRIS dataset, building a logistic regression for multi-
3a 16 7
class classification and testing the model on the IRIS dataset downloadable
from UCI Machine Learning Repository
● Building K-nearest neighbor classifier and testing on the IRIS dataset
downloadable from UCI Machine Learning Repository
(Use Scikit-learn open source data analysis library for implementing the
models)

• Building a naive Bayes classifier for sentiment analysis

3b 4 3
(Use Scikit-learn open source data analysis library)

● Loading csv file based datasets using file read operation in python
3c ● Introduction to pandas library and loading csv and json files 4 3
● Using Scikit-learn library to develop decision tree classifier in python.

4. Unsupervised Learning [ 3 Marks ]

4a Using Scikit-learn library to use the K-means algorithm for clustering IRIS data
4 3
and its visualization

5. Artificial Neural Network [ 6 marks ]

● Using MLP from Scikit learn library, develop a handwritten digit recognition
model using MLP and MNIST dataset
● Using CNN from keras library, develop a handwritten digit recognition
5a 6 6
model using CNN and MNIST dataset
● Compare the performance of the MLP based model and the CNN based
model for the handwritten digit recognition task