IC25401 Machine Learning and Deep Learning – Semester IV – ICE – R-2025

Subject Code & Name: IC25401 – Machine Learning and Deep Learning

Regulation: R-2025

Semester: IV (Fourth Semester)

Branch: B.E. Instrumentation and Control Engineering (ICE)

Credits / L-T-P: 3 Credits | L-T-P: 3-0-0

Course Objectives

• This course aims to provide a strong foundation in the fundamental concepts of machine learning and its core techniques.
• It introduces deep learning architectures such as CNNs, RNNs, and transfer learning for solving complex tasks in classification and sequence modeling.

Full Unit-wise Syllabus

Unit I – Introduction to Machine Learning

Learning algorithms, Maximum likelihood estimation, Building machine learning algorithm, Neural Networks Multilayer Perceptron, Back-propagation algorithm and its variants Stochastic gradient decent, Curse of Dimensionality.

Activity Based Learning: Implement a simple machine learning model (e.g., linear regression or perceptron) using a dataset in Python.

Unit II – Introduction To Deep Learning &Convolution Neural Networks

Machine Learning Vs. Deep Learning, Representation Learning, Width Vs. Depth of Neural Networks, Activation Functions: RELU, LRELU, ERELU, Unsupervised Training of Neural Networks, Restricted Boltzmann Machines, Auto Encoders. CNN Architectural Overview – Motivation - Layers – Filters – Parameter sharing – Regularization, Popular CNN Architectures: ResNet, AlexNet – Case studies.

Activity Based Learning: Implement activation functions (RELU, LRELU, etc.) and visualize their outputs on sample data. Build a simple CNN for image classification

Unit III – Sequence Modelling – Recurrent and Recursive Nets

Recurrent Neural Networks, Bidirectional RNNs – Encoder-decoder sequence to sequence architectures - BPTT for training RNN, Long Short Term Memory Networks; Case studies.

Activity Based Learning: Build a basic RNN for text prediction

Unit IV – Auto Encoders and Deep Generative Models

Deep Belief networks – Boltzmann Machines – Deep Boltzmann Machine - Generative Adversial Networks – Case studies.

Activity Based Learning: Visualize how an autoencoder compresses and reconstructs image data.

Course Outcomes (COs)

• CO1: Explain fundamental concepts of machine learning and deep learning, including neural networks, backpropagation, and optimization techniques.
• CO2: Implement and evaluate basic machine learning and deep learning models, including MLPs, CNNs, and RNNs, for classification and sequence prediction tasks.
• CO3: Analyze and design advanced deep learning architectures such as CNNs, LSTMs, and sequence-to-sequence models for complex applications.
• CO4: Understand and implement autoencoders and deep generative models (e.g., GANs, Boltzmann Machines) for representation learning and data generation.

Assessment Pattern (Quick Note)

• Weightage: Continuous Assessment 40% | End Semester Examinations 60%
• Internal methodology: Assignments (20%), Solution to application-oriented problems using software (20%), Projects based on open source data (like kaggle) (20%), Internal Examinations (40%)

Source: Official Anna University – B.E. Instrumentation and Control Engineering R-2025 Syllabus
Last Updated: October 2026