ARTIFICIAL INTELLIGENCE TOOLS AND APPLICATIONS

(Non CS-IT)
Course Code: 19CS11D1 L T P C
2 0 2 3

Course Outcomes: At the end of the course, a student will be able to:
CO1: Demonstrate various AI applications, languages and Intelligent Agents.
CO2: Solve problems using search strategies and understand the basic process of Machine Learning.
CO3: Apply classification and regression algorithms on real world data.
CO4: Develop an expert system.
CO5: Comprehend the structure of an artificial neural network and identify the building blocks of a
convolutional neural network.

UNIT-I: (6 Lectures)
ARTIFICIAL INTELLIGENCE: Introduction, Definition of AI, Goals of AI, Turing Test, Applications
of AI, AI Programming Languages; Introduction, Intelligent Systems, the Concept of rationality, types of
Agents, Environments and its properties, PEAS. (Textbook 1: Chapters-1,2)
Learning Outcomes: At the end of the unit, student will be able to
1. classify various AI Applications. (L2)
2. list the AI Languages. (L1)
3. explain various types of Agents. (L2)

UNIT-II: (6 Lectures)
SEARCH STRATEGIES: Introduction, Brute Force or Blind Search, Breadth-First Search, Depth-First
Search, Hill Climbing, Best-First Search. (Textbook 1-Chapters-4,5)
MACHINE LEARNING: Introduction, Machine Learning Process, Feature Engineering-Feature
Extraction, Feature Selection, Feature Engineering Methods, Feature Engineering, Data Visualization-
Line Chart, Bar Chart, Pie Chart, Histograms, Scatter Plot, Seaborn-Distplot, joint plot. (Textbook 2-
Chapters-1,3,4)
Learning Outcomes: At the end of the unit, student will be able to
1. apply informed search techniques to problems. (L3)
2. interpret the features using feature engineering. (L2)
3. analyse the data using different visualization techniques. (L4)

UNIT-III: (7 Lectures)
REGRESSION: Simple Regression, Multiple Regression, Model Assessment-Training Error,
Generalized Error, Testing Error, Bias-Variance Tradeoff (Textbook 2-Chapter-5)
CLASSIFICATION: Linear Classification, Logistic Regression, Decision Trees (Textbook 2-Chapter-7)
Learning Outcomes: At the end of the unit, student will be able to
1. analyse different classification models and make recommendations towards learning. (L4)
2. solve real world data using classification techniques. (L3)
3. understand different regression models and about its problems. (L2)
UNIT-IV: (6 Lectures)
CLUSTERING: K-Means Clustering. (Textbook 2-Chapter-8)
EXPERT SYSTEMS: Introduction, Need and Justification of ES, Knowledge Representation,
Knowledge Acquisition and Variation, Utilisation and Functionality, Basics of Prolog. (Textbook - 1-
Chapter-14)
Learning Outcomes: At the end of the unit, student will be able to
1. understand the concept of clustering over classification. (L2)
2. distinguish between expert systems and traditional systems. (L2)
3. identify different applications of expert systems. (L3)

UNIT-V: (7 Lectures)
ARTIFICIAL NEURAL NETWORKS (ANNs): Biological Neuron, Types of ANN, Optimization
Techniques, Vanishing Gradient Problem, Exploding Gradient Problem, Weight Initialization. (Textbook
- 3-Chapter-1)
CONVOLUTION NEURAL NETWORKS(CNNs): Introduction, Components of CNN Architecture-
Convolution Layer(with example), Pooling/Downsampling Layer, Flattening Layer, Fully Connected
Layer; Rectified Linear Unit Layer, Exponential Linear Unit,Unique Properties of CNN, Architectures of
CNNs, Applications of CNN. (Textbook - 3-Chapter-2)
Learning Outcomes: At the end of the unit, student will be able to
1. understand the architecture of an artificial neuron. (L2)
2. illustrate different artificial neural network architecture. (L2)
3. analyse the effect of different activation functions of a CNN unit. (L4)

TEXT BOOKS:
1. Dr.Nilakshi Jain, Artificial Intelligence, As per AICTE: Making a System Intelligent, Wiley
Publications, 1 Edition,2019.
st

2. Vijayvargia, Abhishek, Machine Learning with Python: An Approach to Applied Machine

Learning, BPB Publications; 1 edition,2018.
st

3. Dr.S.Lovelyn Rose, Dr. L.Ashok Kumar, Dr.D.Karthika Renuka, Deep Learning using Python,
Wiley India Pvt. Ltd 2019.

REFERENCES:
1. Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach, Pearson
Publications, 4th Edition, 2020.
2. Saroj Kaushik, Artificial Intelligence, Cengage Learning India, 2011.

WEB REFERENCES:
1. https://keras.io/
2. https://ai.google/
3. https://www.coursera.org/learn/neural-networks-deep-learning#syllabus
4. https://swayam.gov.in/nd1_noc19_me71/preview
LIST OF PROGRAMS:
All the programs should be implemented in Python:

1. Write a program for the following

a. To generate an array of random numbers from a normal distribution for the array of a given
shape.
b. Implement Arithmetic operations on two arrays (perform broadcasting also.)
c. Find minimum, maximum, mean in a given array. ( in both the axes )
d. Implement np.arange and np.linspace functions.
e. Create a pandas series from a given list.
f. Create pandas series with data and index and display the index values.
g. Create a data frame with columns at least 5 observations
i. Select a particular column from the DataFrame
ii. Summarize the data frame and observe the stats of the DataFrame created
iii. Observe the mean and standard deviation of the data frame and print the values.

2. Write a Program to determine the following in the Titanic Survival data.

a. Determine the data type of each column.
b. Find the number of non-null values in each column.
c. Find out the unique values in each categorical column and frequency of each unique value.
d. Find the number of rows where age is greater than the mean age of data.
e. Delete all the rows with missing values.

3. Perform Data Analysis on the Titanic Data Set to answer the following.
a. Information regarding each column of the data
b. Impact of each column on the label
c. Number of survivals in each gender
d. Number of survivals in each passenger class
e. The number of people who are not alone.

4. Perform Data Analysis on the California House Price data to answer the following
a. Data Type of each column and info regarding each column
b. The average age of a house in the data set.
c. Determines top 10 localities with the high difference between income and house value. Also,
top 10 localities that have the lowest difference
d. What is the ratio of bedrooms to total rooms in the data
e. Determine the average price of a house for each type of ocean_proximity.

5. Write a program to perform the following tasks

a. Determine the outliers in each non-categorical column of Titanic Data and remove them.
b. Determine missing values in each column of Titanic data. If missing values account for 30%
of data, then remove the column.
c. If missing values are less than 30% of entire data then create a new data frame
i. Missing values in numeric columns are filled with the mean of the corresponding
column.
 ii. Missing values in categorical columns are filled with the most frequently
occurring value.

6. Write a program to perform the following tasks

a. Determine the categorical columns in Titanic Dataset. Convert Columns with string data type
to numerical data using encoding techniques.
b. Convert data in each numerical column so that it lies in the range [0,1]

7. Implement the following models on Titanic Dataset and determine the values of accuracy,
precision, recall, f1 score and confusion matrix for the test data.
a. Logistic Regression
b. Random Forest Classifier

8. Implement the following models on the California House Pricing Dataset and determine the
values of R score, the area under roc curve and root mean squared error for the test set.
2

a. Linear Regression with Polynomial Features

b. Random Forest Regressor

9. Design an expert system.

(Suggested exercise: for detecting infant is diabetic or not by considering the following
parameters:
a. age = [int from 0 to 5]
b. glycemie= [ int 2/l ]
c. Signs like shakiness, hunger, sweating, headache, diabetic_parents, pale, urination, thirst,
blurred_vision, dry_mouth, smelling_breath, shortness_of_breath = [Boolean: True/False]
Knowledge Engine:
def concerned_person(self)
def hyper_glycemy(self, glycemie)
def hypo_glycemy(self, glycemie)
def has_signs_low_sugar(self, age)
def protocole_risk_low(self)
def protocole_alert_low(self)
def has_diabetic_parents(self)
def has_signs_high_sugar(self, **_)
def protocole_risk_high(self)
def protocole_alert_high(self) )

10. Implement a single neural network and test for different logic gates.

11. Write a program to train and test a Convolutional Neural Network to determine the number, given
an image of a handwritten digit. Determine the training and validation accuracies of your model. (Train
your model for 5 epochs).

SUGGESTED DATASETS:

1. https://www.kaggle.com/c/titanic/data
2. https://www.kaggle.com/camnugent/california-housing-prices
3. https://www.kaggle.com/c/house-prices-advanced-regression-techniques