Indeed, a waste of money and time. And there is no way to predict the future of Bitcoin and I think, we've done with our own market analysis and TA which I think we don't need such machine.
Data collection is never a waste of time or money. Like the Ambanis have said, "Data is the new oil"
I don't know how the invention is made and how it proves that it predicts the right price. We keep thinking about the volatility of the market which makes these predictions and any form of predictions will never be perfect or even too close to it.
See, you dont have to predict, but use an algorithm to collect the data. From that you make predictions on your own and see how much you can get correct. It is possible to get more than 50% correct. But the bottom line is that market moves with a lot of factors that you can possibly never complete to include in the calculation.
Hence these can only be "predictions" and therefore not "facts". Facts are 100% correct while predictions are not.
Goodluck with ML, one splash of a whale and its gone
An outlier value does not splash away the data which has been collected. The power of statistics is tested here but this is not the first time you are getting outlier values, and that is why research uses statistics on a large scale.
