Deep Learning With TensorFlow Training and Certification

Deep learning is the machine learning technique applicable for most new capabilities in areas like natural language processing, robotics, image recognition and artificial intelligence. This chapter provides an introduction to Deep Learning, its advantages and applications and introduction to Machine Learning.

Class 1.1:

• Deep Learning Overview
  
• Machine Learning Introduction
  
• Benefits of Deep learning compared to Machine Learning
  
• Applications of Deep Learning
  

Deep learning algorithms help you identify the unknown structure in the input distribution to discover useful representations. The observations are available often at multiple levels, with higher-level learned features defined in the place of lower-level features

Class 2.1:

• Machine Learning Algorithms
  
• Linear Algebra and Statistics
  
• About Reinforcement learning
  

Multiple layers of interconnected nodes arranged like neurons in the human brain are known as Neural Networks. Initially, Training data is fed into an input layer which is fed into any number of hidden layers. Image processing occurs in the hidden layer, and weighted functions are then applied.

Class 3.1:

• Define Neural Networks
  
• Techniques of Neural Networks
  
• Different Activation and Loss Functions
  
• Different parameters of Neural Networks
  

This chapter explains you about the deeper networks that handle multiple hidden layers. Deeper networks build representations of patterns in the data and replace the requirement for featured learning.

Class 4.1:

• Deep Learning Definition
  
• Architectural Principals of Deep Networks
  
• Different parameters of Deep Networks
  
• Building Blocks of Deep Networks
  
• Use of Reinforcement learning in Deep Network
  

Class 5.1:

• TensorFlow Introduction
  
• Use of TensorFlow in Deep Learning
  
• Working of TensorFlow
  
• TensorFlow Installation
  
• HelloWorld with TensorFlow
  
• Machine learning algorithms on TensorFlow
  

Conventional Neural Networks (CNNs) consists of multiple hidden layers which process the output of the prior layer. CNNs are useful in analyzing visual imagery. The Tensor Flow layers support a high-level API for constructing a neural network. Tensor flow Layers supports the creation of dense layers and convolutional layers

Class 6.1:

• Introduction to CNN
  
• Application of CNNs
  
• Architecture of a CNN
  
• Convolution and Pooling layers in a CNN
  
• Understanding and Visualizing a CNN
  
• Transfer Learning and Fine-tuning Convolutional Neural Networks
  

The neural network involving a sequence of inputs is known as recurrent neural networks. Recurrent neural networks include connections which have loops, add feedback and memory to the networks over time. Memory allows the network to learn and generalize across sequences of inputs than individual patterns.

A dominant type of Recurrent Neural Network called the Long Short-Term Memory Network is available for a deep configuration.

Class 7.1:

• Introduction to RNN
  
• Application of RNNs
  
• Long Short-Term Memory(LSTM)
  
• Recursive Neural Tensor Network Theory
  
• Working with Neural Network Model
  

Restricted Boltzmann machines referred as RBM is active in deep learning networks. In particular, deep belief networks use RBM by stacking them and optionally fine-tuning the resulting deep network with gradient descent and backpropagation.

Class 8.1:

• Applications of RBM
  
• Collaborative Filtering with RBM
  
• Introduction to Autoencoders
  
• Autoencoders applications
  
• Understanding Autoencoders
  
• Variational Autoencoders
  
• Deep Belief Network