Deep Learning

How to train a neural network for training regression model?  In previous class we discus about to train multiclass classification to convert it to regression model we have to just change last neuron to single perceptron 

 How to train a neural network for multiclass classification?  In previous class  we learnt about to Binary Classification  in neural network to change it to multiclass classification we do S1 S1="In neural network we have to just change the last neuron units to no.classes"

 How to train a neural network  for Binary classification ? 1. Install anaconda 2. Open jupyter notebook 3. Create file called app.py 4. Click Here to See Code

 In previous class we learn what is MLP? but we listen a term called gradient descent today we study gradient descent. Gradient descent In mathematics, gradient descent is a first-order iterative optimization algorithm for finding a local minimum of a differentiable function. The idea is to take repeated steps in the opposite direction of the gradient of the function at the current point, because this is the direction of steepest descent. Here it is explained by image

Multilayer Perceptron Multilayer Perceptron is collection of Perceptron ANN(MULTILAYER PERCEPTRON) Backpropagation Backpropagation is a process to find weights and baises Step 1 : Initialize all weight and biases Step 2:Choose a Loss function Step 3: Apply Gradient Descent Technique to find correct weights and biases

 Perceptron is a general thing used in deep learning .You know what is neuron in brain Perceptron Layers 1. Input-In Input we provide Input to our model 2. Calculation-we calculate weight and biases for out Perceptron 3.Output-(sigma of wixi)+b Perceptron Architecture