Unit 1 Fundamentals of Machine Learning

6 lecture hours

• Four branches of machine learning
• Evaluating machine-learning models
• Data preprocessing
• feature engineering, and feature learning
• Overfitting and under fitting
• The universal workflow of machine learning

Unit 2 Artificial Neural Networks and Deep Neural Networks 11 20

• Program Name: Master of Engineering Level: PG Branch: Information Technology Course / Subject Code : ME01000731 Subject Name: Deep Learning A first look at a neural network
• Data representations for neural networks
• the gears of neural networks: tensor operations
• The engine of neural networks: gradient-based optimization
• Anatomy of a neural network
• Introduction to Keras
• Setting up a deep-learning workstation
• Classifying movie reviews: a binary classification example
• Classifying newswires: a multiclass classification example
• Predicting house prices: a regression example

Unit 3 Deep Learning for Computer Vision

11 lecture hours

• Introduction to convenets
• The convolution operation
• The max-pooling operation
• training a convnet from scratch on a small dataset
• the relevance of deep learning for small-data problems
• Downloading the data
• building your network
• Data preprocessing
• Using data augmentation
• Using a pretrained convnet
• Feature extraction
• Fine- tuning
• Visualizing what convnets learn
• Visualizing intermediate activations
• Visualizing convnet filters
• Visualizing heatmaps of class activation

Unit 4 Deep Learning for Text and Sequences

11 lecture hours

• Working with text data
• One-hot encoding of words and characters
• Using word embeddings
• Putting it all together: from raw text to word embeddings
• Understanding recurrent neural networks
• A recurrent layer in Keras
• Understanding the LSTM and GRU layers
• A concrete LSTM example in Keras
• Advanced use of recurrent neural networks
• A temperature-forecasting problem
• Preparing the data
• non-machine-learning baseline
• A basic machine-learning approach
• A first recurrent baseline
• Using recurrent dropout to fight overfitting
• Stacking recurrent layers
• Using bidirectional RNNs

Unit 5 Recent Trends in Deep Learning Architectures, Residual Network, Skip

• Connection Network
• Fully Connected CNN etc. 6 20