Machine learning beats BTC/USDT on unseen data, even with transaction fees and slippage. There are a lot of articles about experimentation applying machine learning to crypto trading but it is hard to find one with realistic methodology. Ideally, results should come either from trading history on an actual exchange or from a simulation with unseen data and included transaction fees.
That’s why I wrote this article — I want to tell you how I do my financial market research, present some of its findings and eventually show you actual results. All with assumptions grounded in reality — including transaction fees, slippage and unseen data.
Source: medium.com