SECOND SEMESTER 2024-2025

COURSE HANDOUT (PART-II)

Date: 06/01/2025

In addition to Part-I (General Handout for all courses appended to the timetable), this portion
gives further specific details regarding the course.

Course Code : ME F321

Name of the Course : Data Mining in Mechanical Sciences
Instructor-In-Charge : Dr. Kundan Kumar Singh

Scope and Objective of the Course: The primary focus of this course is on application of data
mining methods in mechanical sciences and engineering. Following are the details objectives of
this course:
Understanding the fundamental of data mining.
Building mathematical foundation for data mining.
Knowledge of data capturing to minimize the information losses.
Understanding of patterns in data, data classification and clustering.
Knowledge of application of identified patterns in data in fields like in-processing
monitoring, machine monitoring, defect analysis, machine and maintenance scheduling etc.
Understanding of different prediction methods based on the data processing.
Understand the difference between artificial intelligence, machine learning and deep
learning.
Knowledge of which data mining methods should be used for what type of data.
Understanding the concept of big data in manufacturing and how to handle it.

Text Books
[1] Bhatia, P., 2019. Data mining and data warehousing: principles and practical techniques.
Cmbridge University Press.
[2] Zaki, M.J. and Meira Jr, W., 2020. Data Mining and Machine Learning: Fundamental
Concepts and Algorithms. Cambridge University Press
Reference Books
[3] Kantardzic, Mehmed. Data mining: concepts, models, methods, and algorithms. John Wiley
& Sons, 2011.
[4] Hand, David J. "Data mining: statistics and more?" The American Statistician 52, no. 2,
1998.
[5] Maimon, Oded Z., and Lior Rokach. Data mining with decision trees: theory and
applications. Vol. 81. World scientific, 2014.
Course Contents
Lecture Learning Objectives Topic to be covered Chapter in
No. the
text book
1-2 Understanding the Why data analysis? Importance of data Class notes
objectives of data analysis, Data analysis to data mining, and
mining Definition of data mining, [3]-1
Objectives of data mining, Analogy of
data mining with concept of mechanical
science, Steps in data mining/Data
mining process, Concept of data
warehouses, Why data mining fails?
3-4 Measuring the data Data measurement (Sampling method) Class notes
and changing the Structure/ representation of data and
form of raw data Temporal and spatial data [4]-2
(Data preprocessing) Different types of data; Numeric and
categorical data Dimensionality of data.,
Mean, variance, data normalization
methods, data smoothing methods.
5-6 Statistical analysis Point estimation, central limit theory, Class notes
and outlier detection Regression methods, logistical and
regression, Density and proximity based [5]-8, [4]-6
outlier detection, clustering based
method.
7-10 Feature Supervised and unsupervised feature, Class notes
identification/extracti Architecture for feature subset selection, and
on methods Principle component analysis for feature [5]-13, [4]-6
selection, relief algorithm, ranking score,
entropy method, Chimerge technique for
feature discretization
11-15 Different clustering Objective of clustering, methods for Class notes
algorithms and its similarity measure, Agglomerative and
applications hierarchical clustering, partitional and [3]-9, [4]-9
incremental clustering, DBSCAN
algorithm, clustering validity.
16-22 Different General approach for classification Class notes
classification problem, accuracy and error rate, and
algorithms and its Decision tree, design of decision tree, [4]-10, [4]-4
application algorithm for decision tree, K nearest
neighbors.
 23-26 Prediction from data; Learning machine, types of learning Class notes
machine learning methods; supervised and unsupervised and
methods learning, statistical learning theory, [3]-6,7
Support vector machine algorithm,
Neural networks; ANN, RNN etc.
27-28 Case studies Different case studies for application of Class notes,
machine learning in mechanical system, Journal
design and manufacturing. papers

List of Experiments
Week Experiment Experiment details
No.
Basics of python, installing python and setting up python
1 E1
environment; how python works?
Getting started with python programming; running code, input,
2 E2
processing and output, running a script
Data loading and preprocessing with pandas, data processing
3 E3
with NumPy, Creating NumPy arrays
Writing control loop, programming for if and if-elase
4 E4
statements, Design and functions
Binary classification and multilabel classification, linear
5 E5
regression, logistical regression
The data pipeline, principle component analysis Kernel PCA,
6 E6
outlier detection binary classification
7 E7 Feature selection, univariant selection, recursive elimination
K-nearest neighbors, Clustering using python DBSCAN in
8 E8
python
Programming for decision tree and entropy analysis, different
9 E9
examples
10 E10 SVM for classification, SVM for regression, tuning SVM
11 E11 Neural network programming
ANN and deep learning using Keras, different case studies
12 E12
using Keras.
Evaluation Scheme and Schedule
Component Duration Weightage(%) Date & Time Nature of
Component
Mid Sem. Test 90 Min. 20 05/03 4.00 - 5.30PM Close Book

Regular classroom 10 --- Open book

interaction
Quiz – 10 To be announced in class Close Book
Project/case study – 15 –- Open Book
Laboratory - 15 To be announced. Open Book

Comprehension 3 Hrs. 30 07/05AN Close Book

examination

VI. Chamber Consultation Hour: To be announced in the class

VII. Notices concerning the course: All notices concerning the course will be displayed on the
CMS notice board.

VIII. Make-up Policy: Make-up will be permitted only in genuine cases with prior permission.

NOTE: The border cases in final grading will be decided based on mainly classroom
attendance and attentiveness in the classroom.
Academic Honesty and Integrity Policy: Academic honesty and integrity are to be maintained by
all the students throughout the semester and any type of academic dishonesty is not acceptable.

Instructor-In-Charge
ME F321