matekrk/CMLDA

This repository contains the code for a Neural Network's implementation using either the Stochastic Gradient Descent or the Conjugate Gradient Descent as optimization algorithm.

CMLDA

Code repository for the Machine Learning group project for the Master's Degree in Computer Science's course Computational Mathematics for Learning and Data Analysis at the Pisa's University.

Assignment

(M) is a NN with topology and activation function of your choice, provided it is differentiable.

• (A1) is a standard momentum descent approach [references: On the importance of initialization and momentum in deep learning].
• (A2) is an algorithm of the class of Conjugate Gradient methods [references: J. Nocedal, S. Wright, Numerical Optimization, A new conjugate gradient algorithm for training neural networks based on a modified secant equation].