CEI357 Deep and Reinforcement Learning Syllabus:

CEI357 Deep and Reinforcement Learning Syllabus – Anna University Regulation 2021

COURSE OBJECTIVES:

1. Understand complexity of Deep Learning algorithms and their limitations
2. Understand the theoretical foundations, algorithms, methodologies, and applications of neural networks and deep learning.
3. It will help to design and develop an application-specific deep learning models.
4. Be capable of confidently applying common Deep Learning algorithms in practice and implementing their own.
5. Be capable of performing experiments in Deep Learning using real-world data.

UNIT I MACHINE LEARNING BASICS

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.

UNIT II INTRODUCTION TO DEEP LEARNING & ARCHITECTURES

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.

UNIT III CONVOLUTIONAL NEURAL NETWORKS

Architectural Overview – Motivation – Layers – Filters – Parameter sharing – Regularization, Popular CNN Architectures: ResNet, AlexNet.

UNIT IV SEQUENCE MODELLING – RECURRENT AND RECURSIVE NETS

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

UNIT V AUTO ENCODERS AND DEEP GENERATIVE MODELS

Deep Belief networks – Boltzmann Machines – Deep Boltzmann Machine – Generative AdversialNetworks.

TOTAL: 45 PERIODS

SKILL DEVELOPMENT ACTIVITIES (Group Seminar/Mini Project/Assignment/Content Preparation / Quiz/ Surprise Test / Solving GATE questions/ etc)

1 Fundamentals of machine learning
2 Fundamentals of deep learning
3 Realization and understanding of CNN
4 Time series forecasting for data
5 Generating of synthetic images

COURSE OUTCOMES:

Students able to
CO1 Have a good understanding of the fundamental issues and basics of machine learning.
CO2 Ability to differentiate the concept of machine learning with deep learning techniques.
CO3 Understand the concept of CNN and transfer learning techniques, to apply it in the classification problems.
CO4 Learned to use RNN for language modelling and time series prediction.
CO5 Use autoencoder and deep generative models to solve problems with high dimensional data including text, image and speech.

TEXT BOOKS:

1. Ian Goodfellow, Yoshua Bengio and Aaron Courville, “ Deep Learning”, MIT Press, 2017.
2. Josh Patterson, Adam Gibson “Deep Learning: A Practitioner’s Approach”, O’Reilly Media, 2017.

REFERENCES:

1. Umberto Michelucci “Applied Deep Learning. A Case-based Approach to Understanding Deep Neural Networks” Apress, 2018.
2. Kevin P. Murphy “Machine Learning: A Probabilistic Perspective”, The MIT Press, 2012.
3. Ethem Alpaydin,”Introduction to Machine Learning”, MIT Press, Prentice Hall of India, Third Edition 2014.
4 Giancarlo Zaccone, Md. Rezaul Karim, Ahmed Menshawy “Deep Learning with TensorFlow: Explore neural networks with Python”, Packt Publisher, 2017.