This is an excellent article which measure some technical analysis indicators and popular machine learning techniques for day trading. Here is the conclusion:
A big takeaway from this project is that the stock market is a very complex system and to explain its behavior with just historical data is not enough. Machine learning algorithms see it as a random walk or white noise. Fundamental analysis, twitter analysis, news analysis, local/global economy analysis — things like this have the potential to improve predictions.
I find it quite interested that this author says Bollinger Bands could work on most stock
There is a source link of Python as well
Remember I have a Python custom building bot course to use technical analysis.
ThanksFACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!