Skip to content

Latest commit

 

History

History
30 lines (19 loc) · 1005 Bytes

cnn.md

File metadata and controls

30 lines (19 loc) · 1005 Bytes
Raw

Convolutional Neural Network

The Convolutional Neural Network (CNN) is an implementation of a basic convolutional neural network in JavaScript. It performs convolution operations on input data using filters.

Class: CNN

Represents a basic convolutional neural network.

Constructor

Creates a new instance of the CNN class.

new CNN(inputSize, numFilters, filterSize, stride, padding)
  • inputSize (number): The size of the input data (assuming square input).
  • numFilters (number): The number of filters to apply.
  • filterSize (number): The size of each filter (assuming square filter).
  • stride (number): The stride value for the convolution operation.
  • padding (number): The padding value for the convolution operation.

Methods

.convolve(input)

Performs the convolution operation on the input data using the defined filters.

  • input (Array<Array>): The input data to perform convolution on.
  • Returns: An array representing the convolved output.