Neural Network Creator was the main piece of code I made for the final group project of my degree, alongside a written report. It is written in Python and uses live financial data from Yahoo Finance, data on yearly global temperature rises from the Met Office, and data on the light curve of stars from NASA's "LightKurve" Python library to test and train neural networks on. The GUI of the program, which uses Python's Tkinter library, enables users to vary the learning rate, number of iterations, and the structure of the neural network in the program - as well as providing the option to compare this against linear, polynomial, and logarithmic regression.
It is my most complex program to date, and can be downloaded HERE. The final version is fairly large at 141MB - a smaller 55MB pre-release version (which only uses Yahoo Finance's stocks data) can be downloaded HERE.
Attached is the presentation that I delivered for the Physics at Lancaster Annual Conference and Exhibition (The PLACE) on my program.