Deep Learning with Python Training

Curriculum

Topics Covered

• Pre-Training Assessment
  • MCQ based pre-training assessment on Python and Deep Learning.
• Module: 1 Python Refresher
  • Data types Sequences
  • list
  • tuple
  • set
  • dictionary Mapping types Program structure Files and console
  • I/O Conditionals
  • Loops Built-ins
• Python Text Basics
  • Introduction to Python Text Basics
  • Working with Text Files with Python
  • Working with PDFs
  • Regular Expressions
  • Assessment
• Module: 2
  • Introduction to Deep Learning
    • Artificial Intelligence
    • Define Deep Learning
    • Neural Networks
    • Deep Learning Applications
  • Perceptron
    • What is a Perceptron
    • Logic Gates with Perceptrons
    • Activation Functions
    • Sigmoid
    • ReLU
    • Softmax
    • Hyperbolic Functions
• Introduction to TensorFlow
  • Computational Graph
  • Artificial Intelligence
  • Key highlights
  • Creating a Graph
  • Regression example
  • Gradient Descent
  • Saving and Restoring Models
  • Tf.layers API
  • Keras-based networks
  • Tensor Board
• Deep Neural Nets
  • Training Deep Neural Nets
  • Vanishing/Exploding Gradients
  • Xavier Initialization
  • Leaky ReLUs and ELUs
  • Batch Normalization
  • Transfer Learning
  • Unsupervised Pre-training
  • Optimizers
  • Regularization
  • Dropout
• Convolutional Neural Networks
  • Intro to CNNs
  • Convolution Operation
  • Kernel filter
  • Feature Maps
  • Pooling
  • CNN Architecture
  • Implement CNN in TensorFlow
  • CNN Workflow for Keras
  • CNN With Keras
  • CNN on Image Data with Keras
• Recurrent Neural Networks
  • Intro to RNNs
  • Unfolded RNNs
  • Basic RNN Cell
  • Dynamic RNN
  • Training RNNs
  • Time-series predictions
  • LSTM
  • Word Embeddings
  • Seq2Seq Models
  • Implement RNN in TensorFlow
  • Implement RNN with keras
• Autoencoders With Convolution Neural Networks (CNN)
  • Introduction of Autoencoders to Deep Learning
  • Autoencoders With CNN- Tensorflow
  • Autoencoders With CNN- Keras
• Post-Training Assessment
  • MCQ based post-training assessment to gauge current understanding on Deep Learning