Lecture-1 Introduction to Deep Learning

·      A revolution in Artificial Intelligence

·      Limitations of Machine Learning

·      What is Deep Learning?

·      Advantage of Deep Learning over Machine learning

·      Reasons to go for Deep Learning

·      Practical Exercise

  Lecture-2 The Math behind Machine Learning: Linear Algebra

·      Scalars

·      Vectors

·      Matrices

·      Tensors

·      Hyperplanes

·      Practical Exercise

  Lecture-3 The Math Behind Machine Learning: Statistics

·      Probability

·      Conditional Probabilities

·      Posterior Probability

·      Distributions

·      Samples vs Population

·      Resampling Methods

·      Selection Bias

·      Likelihood

·      Practical Exercise

  Lecture-4 Review of Machine Learning
 
·      Regression

·      Classification

·      Clustering

·      Reinforcement Learning

·      Underfitting and Overfitting

·      Optimization

·      Practical Exercise

  Lecture-5 Artificial Neural Networks and Various Methods

·      Artificial neural networks

·      Perceptron learning rule

·      Gradient descent rule

·      Tuning the learning rate

·      Regularization techniques

·      Optimization techniques

·      Stochastic process

·      Vanishing gradients

·      Transfer learning

·      Regression techniques

·      Lasso L1 and Ridge L2

·      Unsupervised pre-training

·      Xavier initialization

·      Practical Exercise

  Lecture-6 Fundamentals of Neural Networks Using Tensorflow
 
·      How Deep Learning Works?

·      Activation Functions

·      Illustrate Perceptron

·      Training a Perceptron

·      Important Parameters of Perceptron

·      What is Tensorflow?

·      Tensorflow code-basics

·      Graph Visualization

·      Constants

·      Placeholders

·      Variables

·      Creating a Model

·      Practical Exercise

  Lecture-7 Deep Dive into Neural Networks Tensorflow

·      Understand limitations of A Single Perceptron

·      Understand Neural Networks in Detail

·      Illustrate Multi-Layer Perceptron

·      Backpropagation – Learning Algorithm

·      Understand Backpropagation – Using Neural Network Example

·      MLP Digit-Classifier using TensorFlow

·      TensorBoard

·      Practical Exercise

  Lecture-8 Master Deep Networks

·      Why Deep Learning?

·      SONAR Dataset Classification

·      What is Deep Learning?

·      Feature Extraction

·      Working of a Deep Network

·      Training using Backpropagation

·      Variants of Gradient Descent

·      Types of Deep Networks

·      Practical Exercise

  Lecture-9 Deep Neural Networks (DNNs)

·      Mapping the human mind with deep neural networks (DNNs)

·      Several building blocks of artificial neural networks (ANNs)

·      The architecture of DNN and its building blocks

·      Reinforcement learning in DNN concepts

·      Parameters

·      Layers

·      Optimization algorithms in DNN

·      Activation functions

·      Practical Exercise

  Lecture-10 Convolution Neural Networks (CNN)
 
·      Introduction to CNNs

·      CNNs Application

·      Architecture of a CNN

·      Convolution and Pooling layers in a CNN

·      Understanding and Visualizing a CNN

·      Transfer Learning

·      Fine-tuning Convolutional Neural Networks

·      Practical Exercise

  Lecture-11 Recurrent Neural Networks (RNN)
 
·      Introduction to RNN Model

·      Application use cases of RNN

·      Modelling sequences

·      Training RNNs with Backpropagation

·      Long Short-Term memory (LSTM)

·      Recursive Neural Tensor Network Theory

·      Recurrent Neural Network Model

·      Practical Exercise

  Lecture-12 Restricted Boltzmann Machine (RBM) & Autoencoders

·      Restricted Boltzmann Machine

·      Applications of RBM

·      Collaborative Filtering with RBM

·      Introduction to Autoencoders

·      Autoencoders applications

·      Understanding Autoencoders

·      Practical Exercise

  Lecture-13 Keras

·      Define Keras

·      How to compose Models in Keras

·      Sequential Composition

·      Functional Composition

·      Predefined Neural Network Layers

·      What is Batch Normalization

·      Saving and Loading a model with Keras

·      Customizing the Training Process

·      Using TensorBoard with Keras

·      Use-Case Implementation with Keras

·      Keras API

·      Practical Exercise

  Lecture-14 TFlearn
 
·      Define TFlearn

·      Composing Models in TFlearn

·      Sequential Composition

·      Functional Composition

·      Predefined Neural Network Layers

·      What is Batch Normalization

·      Saving and Loading a model with TFlearn

·      Customizing the Training Process

·      Using TensorBoard with TFlearn

·      Use-Case Implementation with TFlearn

·      Practical Exercise

  Lecture-15 GPU in Deep Learning

·      GPUs’ introduction

·      How are they different from CPUs?

·      Significance of GPUs in training Deep Learning networks

·      Forward pass and backward pass training techniques

·      GPU constituent with simpler core

·      Concurrent hardware

·      Practical Exercise

  Lecture-16 Deep Learning Applications

·      Image processing

·      Natural Language Processing (NLP)

·      Speech recognition

·      Video analytics

·      Practical Exercise

  Lecture-17 Chatbots

·      Microsoft’s Luis

·      Google API.AI

·      Amazon Lex

·      Open–Closed domain bots

·      Generative model

·      The sequence to sequence model (LSTM)

·      Practical Exercise

  Lecture-18 Time Series Analysis

·      What is Time Series

·      Techniques

·      Applications

·      Components of Time Series

·      Moving average

·      Smoothing techniques

·      Exponential smoothing

·      Univariate time series models

·      Multivariate time series analysis

·      Arima model

·      Time Series in Python

·      Sentiment analysis in Python (Twitter sentiment analysis)

·      Text analysis

·      Practical Exercise

  Case Studies

·      Case Study-1 Creating Application-oriented Analyses Using Tax Data

·      Preparing for the analysis of top incomes

·      Importing and exploring the world's top incomes dataset

·      Analyzing and visualizing the top income data of the India

·      Furthering the analysis of the top income groups of the India

·      Reporting with Jinja2                    

·      Case Study-2 Driving Visual Analyses with Automobile Data

·      Preparing to analyze automobile fuel efficiencies

·      Exploring and describing fuel efficiency data with Python

·      Analyzing automobile fuel efficiency over time with Python

·      Investigating the makes and models of automobiles with Python                       

·      Case Study-3 Working with Social Graphs

·      Preparing to work with social networks in Python

·      Importing networks

·      Exploring subgraphs within a heroic network

·      Finding strong ties

·      Finding key players

·      Exploring characteristics of entire networks

·      Clustering and community detection in social networks

·      Visualizing graphs             

·      Case Study-4 Recommending Movies at Scale

·      Modeling preference expressions

·      Understanding the data

·      Ingesting the movie review data

·      Finding the highest-scoring movies

·      Improving the movie-rating system

·      Measuring the distance between users in the preference space

·      Computing the correlation between users

·      Finding the best critic for a user

·      Predicting movie ratings for users

·      Collaboratively filtering item by item

·      Building a nonnegative matrix factorization model

·      Loading the entire dataset into the memory

·      Dumping the SVD-based model to the disk

·      Training the SVD-based model                

·      Case Study-5 Har vesting and Geo locating Twitter Data

·      Creating a Twitter application

·      Understanding the Twitter API v1.1

·      Determining your Twitter followers and friends

·      Pulling Twitter user profiles

·      Making requests without running afoul of Twitter's rate limits

·      Storing JSON data to the disk

·      Setting up MongoDB for storing the Twitter data

·      Storing user profiles in MongoDB using PyMongo

·      Exploring the geographic information available in profiles

·      Plotting geospatial data in Python            

·      Case Study-6 Optimizing Numerical Code with NumPy & Scipy

·      Understanding the optimization process

·      Identifying common performance bottlenecks in code

·      Reading through the code

·      Profiling Python code with the Unix time function

·      Profiling Python code using built-in Python functions

·      Profiling Python code using IPython's %timeit function

·      Profiling Python code using line_profiler

·      Plucking the low-hanging (optimization) fruit

·      Testing the performance benefits of NumPy

·      Rewriting simple functions with NumPy

·      Optimizing the innermost loop with NumPy