AIML305
Fundamentals of Deep Learning
B.Tech (Artificial Intelligence & Machine Learning) • Semester 5
Overview
The main objective of this course is to develop the understanding of key mathematical principles which are used behind the working of neural networks. Convolution Neural Networks and Recurrent Neural Networks have also been covered in this course. This course also provides the details for usage of Deep Learning for Natural Language Processing.
Course Syllabus
Unit I
Introduction to Deep Learning, Bayesian Learning, Overview of Shallow Machine Learning, Difference between Deep Learning and Shallow Learning, Linear Classifiers, Loss Function and Optimization Techniques - Gradient Descent and batch optimization.
Unit II
Introduction to Neural Network, Biological Neuron, Idea of computational units, McCulloch–Pitts unit and Thresholding logic.
Artificial Neural Networks: Single Layer Neural Network, Multilayer Perceptron, Back Propagation through time.
Architectural Design Issues.
Unit III
Difficulty of training deep neural networks, Activation Function, Evaluating, Improving and Tuning the ANN.
Hyper parameters Vs Parameters, Greedy layer wise training, Recurrent Neural Networks, Long Short-Term Memory, Gated Recurrent Units, Bidirectional LSTMs, Bidirectional RNNs.
Unit IV
Convolutional Neural Networks, Building blocks of CNN, Transfer Learning, Pooling Layers, Convolutional Neural Network Architectures.
Well known case studies: LeNet, AlexNet, VGG-16, ResNet, Inception Net.
Applications in Vision, Speech, and Audio-Video.