Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

Training Neural Networks: the Backpropagation Algorithm

Go to class Write review

Overview

This course dives into how neural networks learn from data. You'll implement loss functions to measure prediction errors, understand the intuition and mechanics of gradient descent, master the backpropagation algorithm to calculate gradients, and use an optimizer to update network weights.

Syllabus

Unit 1: Mean Squared Error Loss

Fix the Mean Squared Error Loss Function
Implementing Mean Squared Error Loss with JavaScript Loops
Implementing MSE Loss with Vectorized Operations
Implementing Batch Processing and MSE Loss Calculation

Unit 2: Gradient Descent Fundamentals

Implementing 1D Gradient Descent Update Rule
Experimenting with Large Learning Rates in Gradient Descent
Fixing Gradient Descent Implementation
Early Stopping in Gradient Descent
Implementing Gradient Descent Update Logic

Unit 3: Backpropagation in Neural Networks

Fixing Activation Derivative Arguments in Backpropagation
Fixing Gradient Calculation in Backpropagation
Implementing Backpropagation Gradients in Dense Layer
Gradient Shape Validation in Backpropagation

Unit 4: Backpropagation in Multi Layer Networks

Implementing MSE Loss Derivative for Backpropagation
Complete MLP Backpropagation Loop
Building a Complete Neural Network with Forward and Backward Pass

Unit 5: Training Neural Networks Efficiently

Implementing SGD Weight Updates for Neural Network Training
Implementing Mini-Batch Data Extraction for SGD Training
Fix the SGD Optimizer Update Rule
Putting It All Together: Your First Training Step
Building the Complete Training Loop with Mini-Batch SGD

Reviews

Start your review of Training Neural Networks: the Backpropagation Algorithm