Knowledge Graph — Coursera Notes › Academic disciplines › Computer Science / Information Technology › Artificial Intelligence

Neural Network

concept · part of Artificial Intelligence

A computational model composed of layers of neurons that learn patterns from data through forward and backward propagation.

Common types include feedforward neural networks (FNNs), convolutional neural networks (CNNs) for spatial data, and recurrent neural networks (RNNs) for sequential data.

A feedforward neural network with two hidden layers (64 and 32 neurons) using ReLU activation and a sigmoid output layer is built for binary classification, compiled with Adam optimizer and binary cross-entropy loss.

Inside Neural Network (7)

• Backpropagation — An algorithm that adjusts weights in a neural network based on the error computed during forward propagation to minimize loss.
• Feedforward Neural Network (FNN) — A neural network where data flows in one direction from input to output, used for classification and regression tasks.
• Generative Adversarial Network (GAN) — A neural network architecture with two competing networks (generator and discriminator) that generate realistic synthetic data.
• Autoencoder — An unsupervised neural network that compresses input into a lower-dimensional representation and reconstructs it, used for dimensionality reduction, anomaly detection, and denoising.
• Hidden layer — The layer(s) of a neural network where learning takes place.
• Input layer — The layer of a neural network that receives the data.
• Output layer — The layer of a neural network that produces the final prediction.

Connections

• Uses Function
• Uses Backpropagation
• Builds on Linear Algebra

This is the text view of an interactive 3D knowledge graph — open this page with JavaScript enabled to explore it visually.