if you have a pattern that is profitable, you can train NN with thousands of this pattern and recognize the pattern when it occurs in out-of-sample data sets. the question is: the author even does not know which pattern is profitable, how can he use computer to recognize winning patterns?
how did the author use image to predict price? in your another post in this thread, the author use BULL bar and Bear bar to train the data set, and predict if the next bar will be bull or bear bar. how could the author use image to predict price? I did not get it. what kind of image to feed into...
according to chinese newspaper, there is still a while before the phase one deal can be signed. trump takes every opportunity to manipulate the market.
but my robot can not read the colors in the front of white house. his tweet called me by shock watching es jumped 20 points in 6 minutes. even companies choose to publish big events after market close or way before market open.
his tweet claimed great news and closer to deal with china, which caused stock to rise sharply in the morning. someone has a big loss due to the tweet. but on weekend, there is no deal coming out. can the trader sue trump?
every time my account capital arises to a round number, there will be a drawdown, even though before that it has been climbing for 2 weeks. anyone trades based on technical analysis of portfolia curve?