akarsh-saxena/Neural-Network-From-Scratch

Implementation of Neural Networks with Python

Neural Network From Scratch

This repository contains my implementation of Neural Networks which I follow from Andrew NG's Deep Learning Specialization course on Coursera.

Content

• Logistic Regression as a Neural Network
  Neural Network having a single layer with only one neuron.
  
  
• Shallow Neural Network
  Neural Network having a single layer with 'n' neurons.
  
  
• Deep Neural Network
  Neural Network having 'L' layers with multiple neurons each.
  
  
• L2 Regularization
  Implementation of L2 Regularization (Ridge Regression).
  
  
• Dropout
  Implementation of Dropout.
  
  
• Gradient Checking
  Gradient checking to check whether the implementation of gradient descent algorithm is correct or not.
  
  
• Exponentially Weighted MA (EWMA)
  Implementation of Exponentially Weighted Moving Average.
  
  
• EWMA in Gradient Descent
  Implementation of Exponentially Weighted Moving Average in Gradient Descent.
  
  
• Mini Batch Gradient Descent
  Implementation of gradient descent with mini batches
  
  
• Gradient Descent With Momentum
  Implementation of Gradient Descent with Momentum.