Master Neural Networks: Build with JavaScript and React

Course Curriculum

Chapter 1: Introduction

Lecture 1: Introduction

Lecture 2: How to approach the lectures

Lecture 3: Few words before start

Chapter 2: Neuron vs Perceptron

Lecture 1: Initial Setup

Lecture 2: Neuron

Lecture 3: Biological neuron vs perceptron

Chapter 3: Classify objects

Lecture 1: Define data

Lecture 2: Define data in code

Lecture 3: Weighted sum

Lecture 4: Change the weight

Lecture 5: Update weights

Lecture 6: Compute sums in code

Lecture 7: Update weights for all inputs

Lecture 8: Update weights in code

Lecture 9: Measure accuracy

Lecture 10: Testing data

Lecture 11: Init weights randomly

Lecture 12: Measure acuracy each epoch

Chapter 4: Mnist Dataset

Lecture 1: Mnist data

Lecture 2: Read bytes

Lecture 3: Read info bytes

Lecture 4: Show label file

Lecture 5: Parse labels out

Lecture 6: Parse out images

Lecture 7: Save testing data

Chapter 5: Frontend in React

Lecture 1: Init react app

Lecture 2: Init home and navigation

Lecture 3: Basic router

Lecture 4: Finish routing

Lecture 5: Load mnist data

Lecture 6: Batch the data

Lecture 7: Display all labels

Lecture 8: Display images

Chapter 6: Real data training

Lecture 1: Save training data

Lecture 2: Process labels and inputs

Lecture 3: Train the perceptron

Lecture 4: Testing accuracy

Lecture 5: Show misclassified data

Lecture 6: Export model

Chapter 7: Prediction on Frontend

Lecture 1: Fetch model on frontend

Lecture 2: Make predictions

Lecture 3: Display prediction visualy

Lecture 4: New image prediction page

Lecture 5: Canvas preparation

Lecture 6: Draw on cavnas

Lecture 7: Get inputs from canvas

Lecture 8: Make prediction from canvas

Lecture 9: Clear canvas and display prediction

Chapter 8: Improving the model

Lecture 1: Adjust pixel values

Lecture 2: Experimenting with training

Lecture 3: Get misclassified data ready

Lecture 4: Send data to server

Lecture 5: Store misclassified data

Lecture 6: Simple perceptron wrap up

Chapter 9: Neural Networks – Forward Propagation

Lecture 1: Mlp introduction

Lecture 2: Mlp Finish Network

Lecture 3: Forward pass hidden activations

Lecture 4: Mlp data in code

Lecture 5: Compute hidden sum in code

Lecture 6: Compute hidden activations in code

Lecture 7: Hidden to output sums math + code

Lecture 8: Softmax explanation + math

Lecture 9: Additional info

Lecture 10: More explanation – recap

Lecture 11: Compute output probabilities

Chapter 10: Neural Networks – Backward Propagation

Lecture 1: Code cleanup

Lecture 2: Calculate output deltas

Lecture 3: Delta hidden neuron 1

Lecture 4: Delta Hidden neuron 2

Lecture 5: Hidden deltas in code

Lecture 6: Gradient of loss math

Lecture 7: Update hidden output weights math

Lecture 8: Update hidden output weights in code

Lecture 9: Weights input hidden math

Lecture 10: Weights input hidden code

Chapter 11: Neural Networks – Model Training

Lecture 1: More training data

Lecture 2: Init weghts randomly

Lecture 3: Loss function

Lecture 4: Measure accuracy of NN

Chapter 12: Neural Networks – Train MNIST Dataset

Lecture 1: Generate mlp data

Lecture 2: Load mlp data

Lecture 3: Encode labels

Lecture 4: Train the mlp model

Lecture 5: Improve logging

Lecture 6: Save mlp model

Lecture 7: Improving mlp model

Chapter 13: Neural Networks – Frontend

Lecture 1: Prepare mlp fronted page