I would like to know how to write a custom gradient for a function which have multiple outputs( or an array). For a simple example, I wrote the following code for y=tan( x @ w + b) with x shape is (2,3) and y shape is (2,2). To compare results, I calculated the operation by usual way and by the
Tag: backpropagation
Compute gradients across two models
Let’s assume that we are building a basic CNN that recognizes pictures of cats and dogs (binary classifier). An example of such CNN can be as follows: Let’s also assume that we want to have the model split into two parts, or two models, called model_0 and model_1. model_0 will handle the input, and model_1 will take model_0 output and
Backpropagation with Momentum using Scikit-Learn
I’m trying to use Scikit-Learn’s Neural Network to classify my dataset using a Backpropagation with Momentum. I need to specify these parameters: Hidden neurons, Hidden layers, Training set, Learning rate and Momentum. I found MLPClassifier in Sklearn.neural_network package. The problem is that this package is part of Scikit-learn V0.18 which is a dev version. Is there a way I could