DEEP LEARNING(AI) TRAINING

PYTHON DEEP LEARNING USING TENSOR FLOW AND KERAS

Introduction to Neural Network

• what is neural network..?
• How neural networks works?
• Gradient descent
• Stochastic Gradient descent
• Perceptron
• Multilayer Perceptron
• BackPropagation

Building Deep learning Environment

• Overview of deep learning
• DL environment setup locally
  • Installing Tensorflow
  • Installing Keras
• Setting up a DL environment in the cloud
  • AWS
  • GCP
• Run Tensorflow program on AWS cloud plateform

Tenserfow Basics

• Placeholders in Tensorflow
  • Defining placeholders
  • Feeding placeholders with data
• Variables,
• Constant
• Computation graph
• Visualize graph with Tensor Board

Activation functions

• What are activation functions?
• Sigmoid function
• Hyperbolic Tangent function
• ReLu -Rectified Linear units
• Softmax function

Training Neural Network for MNIST dataset

• Exploring the MNIST dataset
• Defining the hyperparameters
• Model definition
• Building the training loop
• Overfitting and Underfitting
• Building Inference

LEARNING

Word Representation Using word2vec

• Learning word vectors
  • Loading all dependencies
  • Preparing the text corpus
  • defining our word2vec model
  • Training the model
  • Analyzing the model
• Visualizing the embedding space by plotting the model on tensorboard

Clasifying Images with Convolutional Neural Networks(CNN)

• Introduction to CNN
• Train a simple convolutional neural net
• Pooling layer in CNN
• Building ,training and evaluating our first CNN
• Model performance optimization

Popular CNN Model Architectures

• Introduction to Imagenet
• LeNet architecture
• AlexNet architecture
• VGGNet architecture
• ResNet architecture

Introduction to Recurrent Neural Networks(RNN)

• What are Recurrent Neural Networks (RNNs)?
• Understanding a Recurrent Neuron in Detail
• Long Short-Term Memory(LSTM)
• Back propagation Through Time(BPTT)
• Implementation of RNN in Keras

Sequence-to-Sequence Models for Building Chatbot

HandWritten Digits and letters Classification Using CNN

• Code Implementation
  • Importing all of the dependencies
  • Defining the hyperparameters
  • Building a simple deep neural network
  • Convolution in keras
  • Pooling
  • Dropout technique
  • Data augmentation