Hello! I have some experience trading and investing in a wide variety of markets. I am also a programmer and data scientist. Recently, I wrote a mobile application that forecasts the market for photo chart. The forecast is calculated in what percentage of similar cases during the next 10 trading days there was a rise or fall. It is free and is still in closed testing. It is for you and I invite you to join the test and write your opinion, whatever functions you would like to see in it, what you like and what not. So far, a version has been created only for Android and is available via a link in exchange for a gmail email address (Google’s requirement for closed testing). If you send me your address, I will send a link. I promise not to do any mailing and spam. I hope I do not violate the rules of the forum and do not advertise anything. Best regards, Eduard OCF.
Thank you! The program is currently made at the minimum viable product level (MVP). It contains 170,000 patterns based on 11 US indices in the entire history until 2018, inclusive. In the near future I plan to expand the pattern base to millions. Please note that in the first place it is configured to work with daily charts (candlesticks), 40-50 pieces on one chart. My main goal is to understand how to better develop it. For example, what functionality to add, what financial instruments to work with. Thank you in advance for your opinion.
I have a fundamentally different idea. Essentially, it is time series recognition in a huge array from the past. My application does not have any fractals or technical analysis.
it sounds interesting with all the AI and pattern recognition thing. but then trading more than pattern recognition
when I started trading decades ago, I used thousands of indicators. Now I have zero. anyway go experiment with it.