Artificial Intelligence Course and Training (Using Tensorflow)

Deep Learning Course Content

• Introduction to Deep Learning & Neural Networks

  The domain of machine learning and its implications to the artificial intelligence sector, the advantages of machine learning over other conventional methodologies, introduction to Deep Learning within machine learning, how it differs from all others methods of machine learning, training the system with training data, supervised and unsupervised learning, classification and regression supervised learning, clustering and association unsupervised learning, the algorithms used in these types of learning. Introduction to AI, Introduction to Neural Networks, Supervised Learning with Neural Networks, Concept of Machine Learning, Basics of statistics, probability distributions, hypothesis testing, Hidden Markov Model.

• Multi-layered Neural Networks

  Introduction to Multi Layer Network, Concept of Deep neural networks, Regularization. Multi-layer perceptron, capacity and overfitting, neural network hyperparameters, logic gates, thevariousactivationfunctions in neural networks like Sigmoid, ReLu and Softmax, hyperbolic functions. Backpropagation, convergence, forward propagation, overfitting, hyperparameters.

• Training of neural networks

  The various techniques used in training of artificial neural networks, gradient descent rule, perceptron learning rule, tuning learning rate, stochastic process, optimization techniques, regularization techniques, regression techniques Lasso L1, Ridge L2, vanishing gradients, transfer learning, unsupervised pre-training, Xavier initialization, vanishing gradients.

• Deep Learning Libraries

  How Deep Learning Works, Activation Functions, Illustrate Perceptron, Training a Perceptron, Important Parameters of Perceptron,Multi-layer Perceptron What is Tensorflow, Introduction to TensorFlow open source software library for designing, building and training Deep Learning models, Python Library behind TensorFlow, Tensor Processing Unit (TPU) programmable AI accelerator by Google,Tensorflow code-basics, Graph Visualization, Constants, Placeholders, Variables, Step by Step – Use-Case Implementation, Keras.

• Keras API

  Keras high-level neural network for working on top of TensorFlow, defining complex multi-output models, composing models using Keras, sequential and functional composition, batch normalization, deploying Keras with TensorBoard, neural network training process customization.

• TFLearn API for TensorFLow

  Implementing neural networks using TFLearn API, defining and composing models using TFLearn, deploying TensorBoard with TFLearn.

• DNN: Deep Neural Networks

  Mapping the human mind with Deep Neural Networks, the various building blocks of Artificial Neural Networks, the architecture of DNN, its building blocks, the concept of reinforcement learning in DNN, the various parameters, layers, activation functions and optimization algorithms in DNN.

• CNN: Convolutional Neural Networks

  What is a Convolutional Neural Network, understanding the architecture of CNN, use cases of CNN, what is a pooling layer, how to visualize using CNN, how to fine-tune a Convolutional Neural Network, what is Transfer Learning and understanding Recurrent Neural Networks,feature maps, Kernel filter, pooling, deploying convolutinal neural network in TensorFlow

• RNN: Recurrent Neural Networks

  Intro 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, basic RNN cell, unfolded RNN,  training of RNN, dynamic RNN, time-series predictions.

• GPU in Deep Learning

  Introduction to GPUs and how they differ from CPUs, the importance of GPUs in training Deep Learning Networks, the forward pass and backward pass training technique, the GPU constituent with simpler core and concurrent hardware.

• Autoencoders & Restricted Boltzmann Machine (RBM)

  Introduction to RBM and autoencoders, deploying it for deep neural networks, collaborative filtering using RBM, features of autoencoders, applications of autoencoders.

• Deep Learning applications

  • Image Processing
  • Natural Language Processing
  • Speech Recognition
  • Video Analytics

• Chatbots

  Automated conversation bots using one of the descriptive techniques

  • IBM Watson
  • Google API.AI
  • Microsoft’s Luis
  • Amazon Lex
  • Generative
  • Open-Close Domain Bots
  • Sequence to Sequence model (LSTM).

• AI Deep Learning Projects

  Project 1 : Image recognition with TensorFlow

  Industry : Internet Search

  Problem Statement : Building a robust deep learning model to recognize the right object on the internet depending on the user search for the image.

  Description : In this project you will learn how to build Convolutional Neural Network using Google TensorFlow. You will do visualization of images using training, providing input images, losses and distributions of activations and gradients. You will learn to break each image into manageable tiles and input it to the Convolutional Neural Network for the desired result.

  Highlights :

  • Constructing Convolutional Neural Network using TensorFlow
  • Convolutional, Dense & Pooling layers of CNNs
  • Filtering the images based on user queries.

  Project 2 : Building an AI-based chatbot

  Industry : Ecommerce

  Description : This project involves building the chatbots using Artificial Intelligence and Google TensorFlow.

  Problem Statement : Understanding the customer needs and offering the right services through Artificial Intelligence chatbot. You will learn how to create the right artificial neural network with the right amount of layers to ensure the customer queries are comprehensible to the Artificial Intelligence chatbot. This will help to understand natural language processing, understanding beyond keywords, data parsing and providing the right solutions.

  Highlights:

  • Breaking user queries into components
  • Building neural networks with TensorFlow
  • Natural language processing.

  Project 3 : Ecommerce product recommendation

  Industry : Ecommerce

  Problem Statement : Recommending the right projects to customers by artificial intelligence

  Description : This project involves working with recommender systems to provide the right product recommendation to customers with TensorFlow. You will learn how to use Artificial Intelligence to check for user past buying habits, find out what are the products that go hand-in-hand, and recommend the best products for a particular product.

  Highlights :

  • Building neural networks with TensorFlow
  • Looking at huge amounts of data & gaining insights
  • Building recommendation engine with TensorFlow Graph.