Neural Network

A TypeScript implementation of a basic neural network using perceptrons with configurable activation functions.

Architecture

• PerceptronService: Individual neuron with weighted inputs and activation function
• LayerService: Collection of perceptrons forming a network layer
• OrchestratorService: Manages multiple layers for forward propagation
• Activation Functions: Binary Step and Linear activation implementations

Installation

npm install

Usage

import { PerceptronService } from './src/perceptron/services/PerceptronService';
import { BinaryStepService } from './src/activation/services/BinaryStepService';

const activation = new BinaryStepService(0.5);
const perceptron = new PerceptronService(activation);
perceptron.setWeights([0.8, 0.4]);
const result = perceptron.predict([1, 0]);

Development

npm run test          # Run tests
npm run test-coverage # Run tests with coverage
npm run lint          # Lint code
npm run format        # Format code