BCSE332L - Deep Learning

Instructor: Dr. Bhargavi R

Slot And Venue Details

Course Overview

This course provides a comprehensive foundation in deep learning, guiding students from the fundamentals of neural networks through to the state-of-the-art architectures shaping modern AI systems. Students begin with the core building blocks — activation functions, loss functions, gradient descent, and forward/backward propagation — before progressing to deep convolutional models (ResNet, AlexNet, InceptionNet) for computer vision tasks. The course then covers essential techniques for training deep networks effectively, including mini-batch gradient descent, momentum, RMSProp and Adam optimization, batch normalization, and strategies to address underfitting and overfitting. Students explore sequence modeling through recurrent and recursive architectures — including LSTMs, bidirectional RNNs, and BERT — before advancing to specialized topics such as transfer learning, Generative Adversarial Networks, and object detection frameworks (Region-based CNN, Fast R-CNN, YOLO, SSD). The course concludes with an introduction to deep reinforcement learning, covering Q-learning, policy gradients, and actor-critic methods. More details on the topics covered can be obtained from the syllabus.

Syllabus

You can find the syllabus of this course here

Prerequisites

Prior knowledge of the following subjects help you to understand appreciate the Deep Learning Course better

Linear Algebra
Probability and Statistics
Calculus
Machine Learning
Python

Textbooks

Following are the text books for reference -

Ian Goodfellow Yoshua Bengio Aaron Courville, Deep Learning, MIT Press, 2017.
Michael Nielsen, Neural Networks and Deep Learning, Determination Press, first Edition, 2013.
N D Lewis, Deep Learning Step by Step with Python, 2016.
Josh Patterson, Adam Gibson, Deep Learning: A Practitioner's Approach, O'Reilly Media, 2017
Umberto Michelucci, Applied Deep Learning. A Case-based Approach to Understanding Deep Neural Networks, Apress, 2018.
Giancarlo Zaccone, Md. RezaulKarim, Ahmed Menshawy, Deep Learning with TensorFlow: Explore neural networks with Python, Packt Publisher, 2017

Tentative Schedule

Date	Lecture	Readings	Announcements
Welcome to Deep Learning!
Tue, 7th July	Lecture 1: Meet and Greet, Introduction to Deep Learning [Slides ]	Ch 1 - Ian Goodfellow	Model Question Paper Question Paper
Wed, 8th July	Lecture 2: ML paradigms,Loss functions, Model eavaluation, Validation [Slides ]	Ch1, Ch5 - ISLR
Fri, 10th July	Lecture 3: Introduction to ANN, History of ANN [Slides ]	Ch 1 - Ian Goodfellow
Tue, 14th July	Lecture 4: Perceptron [Slides ] [PyTorch Introduction ]	Ch 1 - Michael Nielsen
Wed, 15th July	Lecture 5: Gradient descent [Slides ]	Ch 4 - Ian Goodfellow
Fri, 17th July	Lecture 6: Gradient discent, Learning in MLP [Slides ]	Ch 1 - Ian Goodfellow Ch 2 - Michael Nielsen
Tue, 21st July	Lecture 7: Backpropagation [Slides ]	Ch 1 - Ian Goodfellow Ch 2 - Michael Nielsen
Wed, 22nd July	Lecture 8: Activation functions [Slides ]	Ch 6 - Ian Goodfellow
Fri, 24 July	Lecture 9: Parameters & Hyperparameters, Bias & Variance [Slides ]	Ch 7 - Ian Goodfellow
Tue, 28 July	Lecture 10: Regularization, Ensenbled models, L1, L2 regularization [Slides ]	Ch 7 - Ian Goodfellow
Wed, 29 July	Lecture 11: Early stopping, BatchNorm, SGD with momentum [Slides ]	Ch 7 - Ian Goodfellow Ch 8 - Ian Goodfellow
Fri, 31 July	Lecture 12: NAG with momentum, AdaGrad [Slides ] [PyTorch - Iris Classification using FFNN ]	Ch 8 - Ian Goodfellow
Tue, 04 Aug	Lecture 13: RMSProp, AdaDelta, Adam [Slides ]	Ch 8 - Ian Goodfellow
Wed, 05 Aug	Lecture 14: Solving numericals and discussion
Fri, 08 Aug	Lecture 15: Solving numericals and discussion
Tue, 18 Aug	Lecture 16: Introduction to computer vision, CNN [Slides ]	Ch 9 - Ian Goodfellow
Wed, 19 Aug	Lecture 17: Pooling, Stride, Padding [Slides ]	Ch 9 - Ian Goodfellow
Fri, 21 Aug	Lecture 18: Transfer Learning, LeNet, AlexNet [Slides ] [AlexNetPaper ]	Research Papers
Tue, 25 Aug	Lecture 19: VGGNet [Slides ] [VGGNetPaper ]	Research Paper
Fri, 28 Aug	Lecture 20: ResNet [Slides ] [ResNetPaper ]	Research Paper
Tue, 01 Sep	Lecture 21: Inception/GoogleNet, Parameter computation problems [Slides ] [GoogleNetPaper ] [Classification using nnModule and GPU ]	Research Paper
Wed, 02 Sep	Lecture 22: Object Detection introduction to basic concepts, Object localization, Sliding window algorithm [Slides ] [D2L Chapter 14 ] [Iris classification with CNN - lab ]
Tue, 8th Sept	Lecture 23: IoU, NMS, Introducto RCNN [Slides ] [RCNN Paper ] [DNN for Vision Chapter ]	Research Paper
Wed, 9 Sept	Lecture 24: RCNN,Fast RCNN [Slides ] [Transfer Learning - lab ]
Fri, 11 Sept	Lecture 25: FC to CNN, YOLO [Slides ] [Fast RCNN Paper ] [DNN for Vision Chapter ]	Fast RCNN Research Paper
Tue, 15 Sept	Lecture 26: YOLO, SSD [Slides ] [YOLO Paper ] [DNN for Vision Chapter ]	YOLO Research Paper
Wed, 16 Sept	Lecture 27: Autoencoder [Slides ]	Ch 14 - Ian Goodfellow
Fri, 18 Sept	Lecture 28 Denoising Autoencoder, VAE, Numericals [Slides ]	Ch 14 - Ian Goodfellow
Tue, 06th October	Lecture 28: GAN [Slides ] [GAN Paper ] [DNN for Vision Chapter ]	Ch 20 - Ian Goodfellow GAN Research Paper