I am currently analyzing the best approach for building a software system for prediction stocks to buy and sell. I am not talking about day trading but short trades. This software would utilize a database that would contain historical data of all stocks in a market containing open, close, and others. I have looked at articles along with source code on genetic algorithms and volume weighted average price (VWAP). What I am hoping you can provide me is a high level step by step approach so I can get my hands around the approach I will take. Here is my question. What is the best approach I should take to analysis historical stock market data to return a list of stocks I should buy? This approach will also include the best approach to take to sell stocks I own. What I am asking for could include genetic algorithms, VWAP, or both, or other algorithms.