Deep Learning

Deep Learning course is offered by Velocity Corporate Training Center. Our mission is to build an ideal institute that radiates Peace and harmony by imparting pure education to all the strata of society. Only professionals can make you professional, Come to us and learn to Earn.

Syllabus:

• Module 1: Introduction to the Program
•  Introduction to Deep Learning
•  Difference between ML and DL
•  What to expect from this program
• Module 2: Structure and learning process in Neural Network
•  what is perceptron
•  what is neural network
•  structure of neural network
•  how data flow in neural network
•  how to control shape of the layer
•  how neural network learn
• Module 3: Types of function use in Neural Network
•  Optimization function o Batch Gradient Decent o Stochastic Gradient Decent o Mini Batch Stochastic Gradient Decent o Mini Batch Stochastic Gradient Decent with Momentum o Ada Grad o Ada Delta o Adam  Activation function o Sigmoid o Tanh o Softmax o Relu o Leaky Relu o P-Relu o Elu o Swish  Loss function o MSE o MAE o MSLE o Binary Cross Entropy o Hinge Loss o Categorical Cross entropy loss function o Sparse categorical cross entropy loss function  What is initializer and its important
• Module 4: Implementation of ANN
• How to code Neural Network
•  How to avoid overfitting - Early Stopping, Drop out etc.
•  Hyperparameter Tunning of ANN
• Module 5: Convolutional Neural Network (CNN)
•  Introduction to CNN
• How image get generated
•  RGB Scale/Gray Scale
•  Difference between ANN and CNN
•  Convolutional Layer
•  What is Stride?
•  What is Zero Padding?
•  Pooling layer o Max Pooling o Avg Pooling o Sum Pooling
•  Implementation of CNN
• Module 6: Transfer Learning and Pretrain model
•  What is transfer learning  How to use transfer learning to classify the images
•  Different way to customized pretrain model o Feature Extraction o Fine tunning 
• Steps need to follow in transfer learning
•  Famous pretrain model available for image classification o VGG16 o Inception (Google net) o ResNet50 o Efficient Net
• Implementation of VGG16