Deep Learning Training in Mumbai
Deep Learning is one of the most exciting aspects of artificial intelligence and machine learning technologies. This open source software library is being developed by google mainly for the sake of machine learning neural network research. There are several powerful tools that are being used in deep network to implement neural networks. Candidates who take up this course become proficient in topics such as Tensor Flow, Neural Networks, ANN, Perceptron and Convolutional Neural Network.
Course Contents
Introduction to Deep Learning
- What is deep learning
- Defining deep learning
- What are neural networks
- Types of deep learning applications
Perceptron
- Introduction to perceptron
- Logic gates with perceptron
- Functions for activation
- Sigmoid and Softmax
- ReLU
- Hyperbolic functions
Training ANNs
- Introduction to ANNs
- Perceptron learning rule
- Gradient descent rule
- Minimizing cost functions
- Tuning learning rate
- Stochastic gradient
- Batch gradient
Multi-Layer ANN
- What is MLP
- Defining MLP
- Forward propagation
- Cost function minimizing
- Backpropagation
- Neural set convergence
- Overfitting
- Hyper Parameters in ANN
- Capacity
Tensor Flow
- What is Tensor Flow
- Introduction to Tensor Flow concepts
- Highlights of Tensor Flow
- Regression in Tensor Flow
- Gradient Descent
- Computational Graph
- Keras based networks
- Tensor Board
- Saving and restoring models
Deep Neural Network
- What are deep neural nets
- Xavier initialization
- Batch normalization
- Naive Bayes
- Transfer
- Unsupervised pre-training
- Regularization and dropout
- RELU and ELU
Recurrent Neural network
- What are RNN
- Defining RNN
- Unfolded RNN
- RNN Cell and basics
- Dynamic RNN and training
- LSTM
- Word embedding
- RNN implementation in Tensor Flow
Convolutional Neural Network
- What are CNNs
- What is a Convolution operation
- Kernel filter and feature maps
- CNN architecture
- CNN and Tensor flow integration
- Pooling
