Number Classifier - Feed Forward Neural Network

This is a simple number classification (multi-layer perceptron / feed-forward network) app built with Python and tkinter. It allows you to train a model to classify numbers based on a reference number and predict whether a given number is higher or lower than the reference. And a somewhat nice GUI. We all like GUI's and some colors, right?😎

Requirements

• Everyone can play around with this app. NO coding skills required😇
• Python 3.x
• Git for Desktop

Installation

1. Navigate to the directory of your choice, click into the address bar, type "cmd" and hit Enter.
2. Clone the repository with the command: git clone https://github.com/df8819/SimpleNN.git in the CMD Window.
3. Open the cloned directory/folder.
4. Double click: "Update.bat".
5. For later instances, use: "Run.bat"

Features

• Reference Number: Enter a decimal number between 0 and 1. This number serves as a reference for the classification.
• Number to predict: Enter a decimal number between 0 and 1. This number will be classified as higher or lower than the reference number.
• Layers/Nodes: Enter the number of layers and nodes to use in the neural network.
• Random number count: Enter the count of random numbers to generate for training the model.
• 'Predict' Button: Click this button to perform the prediction based on the provided inputs.
• 'Reset Graph' Button: Click this button to clear the training progress graph.
• 'Visualize Brain' Button: Opens a window for a visual representation of the current model.
• 'Training Data' Button: Opens a window for a visual representation of the random numbers used to train the model. The 'Digit Distribution' Button is a random experimental feature👀
• Note: The prediction will be displayed as % value after the training.

Guide

• A simple user guide is implemented in the app itself. Click the 'Guide' button for more information.

Note:

• The Number Classifier application is a simplified example and may not provide accurate predictions for complex scenarios. It is intended for learning purposes and as a starting point for further development and a better understanding of neural networks.