Steam Market Analyser

Languages used
 This was written in Python and uses the MATLAB engine API to also run MATLAB code from Python.

What it is
 This is a program for finding video game items on the Steam Market that are consistently increasing in price.


How it works
 The price histories of items are gathered from the steam website using the requests library for python. MATLAB code then finds the lines of best fit of the price histories using linear regression. Items that have a large positive slope and strong correlation are considered possible good investments. The MATLAB engine is also used to plot price history information to provide a way for users to rapidly check recommendations.

Potential improvements
 Virtually all items either have turbulent prices or are consistently decreasing. This system could be improved by first ruling out consistently decreasing items using lines of best fit. The remaining items could be checked with a neural network try to guess good recommendations using variables like price history, game the item belongs to and change in game’s Steam players per day. I think this has a good chance of working since people have reported success using neural networks to predict when stocks will be rising or falling in value.