In this notebook, I used Ta-lib's abstract API to add all the financial indicators that Ta-lib has to offer. When dealing with a dataset that contains multiple instruments or stocks, one needs to avoid applying the indicators to the entire dataset all-in-once; the correct way is to calculate each stock's indicators individually. This notebook works for datasets that has only one stock or multiple stocks.The indicator is filtered by finding out the corrolation between this term's indicator result and the next term's return rate. The more positive the corrolation is the stronger the indicator's ability to perdict the next term's return is.