How to Value/Evaluate Algo/Mining Trading Software
Reda Wahba, Co-Founder at QuantMining
Monday, October 2, 2017
Hi Everyone,
Appreciate your help and assistance for an advice as I am new to this group and the whole trading world,
I have a valuable experience on the data mining / machine learning areas as I worked over 15 Years on IBM as a software engineer most of them on this technology, as a matter of fact IBM developed a more advanced search engine based on this technology way earlier than google/yahoo but was specific for IBM main frame customers, that’s why It didn’t show up for public/rest of the world,
Anyway, by using data mining/ML techniques but this time for financial data, a patterns could be extracted to form rule sets that can give trading recommendations based on lots of factors/attributes/. etc.
For the last two years I have gathered a public historical data for over 12000 different market instruments (currencies, commodities, indices, stock exchanges as NYSE, NASDAQ, S&P, JP, HK, etc.) summing all together over 130,000 historical years data, and developed a mining techniques by applying them a yearly profit generated with an average of 40% and sometime way more,
P.S. all historical data gathered is less than 10% of amount of actual historical data in the world, and with more data a more robust rule sets generated. (as all data collected so far only covers 42% of mining rule sets covering all trading cases, and with more data a close to perfect rule sets generates.
I need you as experienced professionals in this area to give me your opinions, is it meaningful, useful?
And if so, how can I reach investment banks/hedge/partners that can take this work to next step??
Appreciate your help,