GUJARAT TECHNOLOGICAL UNIVERSITY

Syllabus for Master of Computer Applications, 3rd Semester With effective

Subject Name: Machine Learning from academic
Subject Code: 639402 year 2020-21

Prerequisites: Basics of computer science including algorithms, data structure, Probability

theory and Bayesian Concept Learning *

1. Teaching and Examination Scheme:

Teaching
Credits Examination Marks
Scheme Total
Theory Marks Practical Marks Marks
L T P C
ESE (E) PA (M) ESE (V) PA (I)
3 - 2 4 70 30 30 20 150
1. Course Outcomes:

Course Outcome Course Outcome (Learner will be able to)

Component
CO-1 Understanding Various Machine Learning Algorithms & Their Area
of Applications
CO-2 Explain the Process of Formulating & Solving Real World Problem
using Machine Technology
CO-3 Design and Implement various Decision Making Problems using
Machine Learning

2. Course Duration: The course duration is of 40 sessions of 60 minutes each.

3. Course Contents:
No. of 70 Marks
Unit
Contents Session (External
No:
s Exam)
 Introduction to Machine Learning:
What is human learning? What is Machine Learning?
I Human learning versus machine learning, Types of 4
machine learning, Applications of machine learning,
Tools for machine learning.
 Preparing to Model , Feature Engineering:
Machine Learning activities, Basic Types of data in
Machine Learning, Structures of data, Data Quality and
II 10
Remediation, Data Pre-Processing,
Introduction to Feature Engineering, Feature
Transformation, Feature Subset Selection
 Modelling and Evaluation:
Selecting a Model, Training a Model, Model
III Representation and Interpretability, Evaluating 8
Performance of a Model, Improving Performance of a
Model
 Supervised Learning: Classification and Regression:
IV Example of Supervised Learning, Classification Model, 10
Classification Learning Steps, Common Classification

Page no. 1 of 6
 GUJARAT TECHNOLOGICAL UNIVERSITY
Syllabus for Master of Computer Applications, 3rd Semester With effective
Subject Name: Machine Learning from academic
Subject Code: 639402 year 2020-21

Algorithms
Example of Regression, Common regression Algorithms
 Unsupervised Learning:
Unsupervised Learning versus Supervised Learning,
V Applications of 8
Unsupervised Learning, Clustering, Finding Pattern using
Association Rules
* Additional Sessions may be taken to revise Prerequisites Contents.
4. Pedagogy:
 ICT enabled Classroom teaching
 Case study
 Practical to be implemented using python library
(Sample practical list is attached as Appendix-A)
 Interactive class room discussions

5. Evaluation:
Students shall be evaluated on the following components:
Internal Evaluation (Internal Assessment- 20)
A
 Continuous Evaluation Component 10 marks
 Class Presence & Participation 10 marks
B Mid-Semester examination (Internal Assessment- 30 Marks)
C End –Semester Examination (External Assessment-70 Marks)

6. Reference Books:
No. Author Name of the Book Publisher
Saikat Dutt, Subramanian
1 Machine Learning Pearson Education
Chandramouli, Amit Kumar Das **
2 Tom M Mitchell Machine Learning McGraw Hill
Anuradha Srinivarasaraghavan,
3 Machine Learning Wiley India
Vincy Joseph
Machine Learning in
4 Peter Harrington DreamTech
Action
Henrik Brink, Joseph Richards, Real-World Machine
5 DreamTech
Mark Fetherolf Learning
Pattern Recognition and
6 Christopher Bishop Springer
Machine Learning
Data Mining: Concepts and
7 Jiawei Han and Michelline Kamber Morgan Kaufmann
Techniques
** The first Reference book is to be considered as Main Book as per Syllabus Contents

Appendix-A

Sample Practical List

1. Write a python code to implement decision tree for below given dataset. Identify the
root node and all subpart or children of node and draw the tree.
Item no Age Income Student Credit Buys-
Rating Computer
1 Youth High No Fair No

Page no. 2 of 6
 GUJARAT TECHNOLOGICAL UNIVERSITY
Syllabus for Master of Computer Applications, 3rd Semester With effective
Subject Name: Machine Learning from academic
Subject Code: 639402 year 2020-21

2 Youth High No Excellent No

3 Middle High No Fair Yes
4 Senior Medium No Fair Yes
5 Senior Low Yes Fair Yes
6 Middle Low Yes Excellent No
7 Senior Low Yes Excellent Yes
8 Youth Medium No Fair No
9 Youth Low Yes Fair Yes
10 Senior Medium Yes Fair Yes
11 Youth Medium Yes Excellent Yes
12 Middle Medium No Excellent Yes
13 Middle High Yes Fair Yes
14 Senior Medium No Excellent No

2. Write a python code to implement K-nearest neighbourhood program for the given
dataset.
Item no Temp Humidity Wind Play
Speed
1 85 85 12 No
2 80 90 9 No
3 83 86 4 Yes
4 70 96 3 Yes
5 68 80 5 Yes
6 65 70 20 No
7 64 65 2 Yes
8 72 95 12 No
9 69 70 5 Yes
10 75 80 2 Yes
11 75 70 3 Yes
12 72 90 4 Yes
13 81 75 5 Yes
14 71 91 15 No

3. Write a python code to implement Apriori algorithm, apply join and prune method and
find frequent item set

Sr# Item Na
no me
1 T1 Bread, butter, milk, soda
2 T2 Coke, egg, milk
3 T3 Bread, butter, egg
4 T4 Break, coke, jam
5 T5 Bread, butter
6 T6 Potato chips, soda

Page no. 3 of 6
 GUJARAT TECHNOLOGICAL UNIVERSITY
Syllabus for Master of Computer Applications, 3rd Semester With effective
Subject Name: Machine Learning from academic
Subject Code: 639402 year 2020-21

7 T7 Coke, fruit, juice

8 T8 Bread, coke, milk
9 T9 Coke, soda, jam, milk
10 T10 Bread, butter, egg, milk,
soda
11 T11 Bread, milk
12 T12 Bread, jam

4. Write a python code to apply Naive Bayesian and Logistic Regression algorithm to
classify that whether a person can buy computer or not based on given test data:

Item no Age Income Student Credit Buys-

Rating Computer
1 Youth High No Fair No
2 Youth High No Excellent No
3 Middle High No Fair Yes
4 Senior Medium No Fair Yes
5 Senior Low Yes Fair Yes
6 Middle Low Yes Excellent No
7 Senior Low Yes Excellent Yes
8 Youth Medium No Fair No
9 Youth Low Yes Fair Yes
10 Senior Medium Yes Fair Yes
11 Youth Medium Yes Excellent Yes
12 Middle Medium No Excellent Yes
13 Middle High Yes Fair Yes
14 Senior Medium No Excellent No

Test Data
Age: Youth Income: LOW Student: No Credit Rating: Fair Buy Computer -??

5. Python code for Preparing to Model (Appendix B.2, Page No. 365 – 373)
and Feature Engineering (Appendix B.4, Page No. 378 – 385).

6. Apply k-means clustering approach with k = 2 to the following dataset.

X Y Z
-0.154 0.376 0.099
-0.103 0.476 -0.027
0.228 0.036 -0.251
0.330 0.013 -0.251
-0.114 0.482 0.014
0.295 0.084 -0.297
0.262 0.042 -0.304

Page no. 4 of 6
 GUJARAT TECHNOLOGICAL UNIVERSITY
Syllabus for Master of Computer Applications, 3rd Semester With effective
Subject Name: Machine Learning from academic
Subject Code: 639402 year 2020-21

-0.051 0.416 -0.306

7. Implement a python program that takes interest rate (x), finds the equation that best
fits the data and is able to forecast out median home price for given interest rate using
the data given below. (Use linear regression)
Interest rate Median home
(%)(x) price (y)
10.3 $183,800
10.3 $183,200
10.1 $174,900
9.3 $173,500
8.4 $172,900
7.3 $173,200
8.4 $173,200
7.9 $169,700
7.6 $174,500
7.6 $177,900
6.9 $188,100
7.4 $203,200
8.1 $230,200
7 $258,200
6.5 $309,800
5.8 $329,800

8. Apply following supervised machine learning algorithms for All classification

problems as shown under:
Algorithm Data set for Classification Problem
Decision Tree Iris Dataset (from sklearn.dataset import load_iris)
Logistic Regression Wine Dataset (from sklearn.dataset import load_wine)
K-Nearest Breast Cancer Wisconsin Dataset (from
Neighborhood sklearn.dataset import load_breast_cancer)
Naïve Bayes Titanic Dataset (https://www.openml.org/d/42438)
SVM Bank marketing Dataset
(https://www.openml.org/d/1461)

Obtain model accuracy,generate classification report, and show the performance of the
model graphically through roc_auc curve.
9. Build a prediction model using regression technique for (1) Boston house-prices (from
sklearn.dataset import load_boston) (2) Diabetes (from sklearn.dataset import
load_diabetes) datasets. Also, evaluate the model.
10. Implement support vector machine approach to predictive modelling for (1) Boston
house-prices (from sklearn.dataset import load_boston) (2) Diabetes (from
sklearn.dataset import load_diabetes) datasets. Also, evaluate the model.
11. Implement unsupervised machine learning algorithm (Clustering – K Means) in
python on Titanic dataset (https://www.openml.org/d/42438) to cluster data by
removing the classlabel.Use elbow method to find the optimal number of clusters.

Page no. 5 of 6
 GUJARAT TECHNOLOGICAL UNIVERSITY
Syllabus for Master of Computer Applications, 3rd Semester With effective
Subject Name: Machine Learning from academic
Subject Code: 639402 year 2020-21

Also, measure the quality of clustering via silhouette coefficient.

12. Implement unsupervised machine learning algorithm (Clustering – K Means) in
python on Breast Tumour dataset to cluster data (use Breast Tumour dataset) by
removing the classlabel.
13. Implement unsupervised machine learning algorithm (Clustering – Hierarchical) in
python on Titanic dataset to cluster data (use Titanicdataset).
14. Implement Apriori algorithm in python to find rules which explain association
between different products for given transactions at a retail store. (The data is
available
athttps://drive.google.com/file/d/1NUXoptUlHY8z4KcFKpFA6sQN5KnWzk3p/view
?us p=sharing)

Desirable Practical Lists

1. Classification and Prediction algorithms on UCI dataset using Python's scikit-learn
library. Perform Exploratory Data Analysis, implement data visualization techniques,
pre-process the data, build a model by applying suitable algorithm, evaluate the model
by generating report, and show the performance of the model graphically through
roc_auc curve.
2. For the sentiment analysis dataset given in link:
https://drive.google.com/file/d/1x6H7_KJjkbDrpgZFS7I2wjsZsILeSJ4S/view?usp=sh
aring, implement the following in python,
a) Clean and pre-process the dataset by removing URL, removing HTML tags,
handling negation words which are split into two parts, converting the words to
lower cases, removing all non-letter characters
b) Split the dataset into training and testing set
c) Implement feature extraction technique (to convert textual data to the numeric
form)
d) Build the classification model using Logistic Regression that classifies if a
given sentiment text is positive or negative
e) Obtain the accuracy score of the built model.
3. Implement a content-based recommender system in python that recommends movies
that are similar to a particular movie using movielens-20m-dataset available at
https://kaggle.com.

The practical exercises should be performed in python.

References:

1) Peter Harrington, “Machine Learning in Action”,DreamTech

2) Michael Bowles, “Machine Learning in Python”,Wiley
3) Gavin Hackeling, Mastering Machine Learning with scikit-learn,Packt
4) Giuseppe Bonaccorso, Machine Learning Algorithms - Second Edition,Packt

Page no. 6 of 6