This project detects trained letters with perceptron algorithm. Initial values, input data and test folder are prepared for Laurene Fausett, "Fundamentals Of Neural Networks" pp.71-76 character recognition example 2.14.
Python 3, Tkinter
python letter_detection.py
Load Button: Training file is loading into view
Save Button: Saving file is saving to file. (Attention: Files first character must be target letter)
Clear Button: Clears the view
Learning Rate: Perceptron learnign rate value. It must be 0 < LR < 1
Threshold: Perceptron threshold value
Max. Iterations: Perceptron maximum iterations value
Train Button: Training folder (data folder)
Weight & Bias Button: Show weights and bias values
Grid: Mouse click swaps on/off neurons
Test Button: Testing current view
Result: Current testing view result
Test Folder Button: Test folder view