The "Multilayer Perceptron" is a modern app written in C++. The application represents a simple artificial neural network in the form of a perceptron, which can be trained on an open dataset and perform recognition of 26 handwritten letters of the Latin alphabet.

make run

• GUI implementation, based on QT6

  

• Load train and test datasets from a csv file.

• Choose the network topology with 2-5 hidden layers.

• Training with using the backpropagation method and sigmoid activation.

• Matrix form: all layers are represented as weight matrices.

• Graph form: each neuron is represented as some node object connected to other nodes by refs.

• Perform experiments on a selected portion of the test sample, defined by a floating-point number ranging from 0 to 1.

• Load BMP images (image size can be up to 512x512) with Latin letters and classify them.

• Draw two-color square images by hand and classify them.

• Real-time training process for a user-defined number of epochs with displaying the error values for each training epoch.

• Run the training process using cross-validation for a given number of groups k.

• Save to a file and load weights of perceptron from a file.

  

Copyright (c). All rights reserved.