Quote from: memehunter on February 19, 2025, 03:25:41 PM

Quote from: xinvisionq on February 19, 2025, 01:54:36 PM

......

So price of bitcoin older than 4 weeks has no impact on current price but principle of superposition does.
You are using the term superposition like some spiritual guru is using by giving the example of Schrödinger's cat.
Again I am not trying to be someone who is just passing some unnecessary sceptical comments. I have gone through your paper and this is my honest opinion. You started your paper with the absolute importance of data in order to make prediction models and when I pointed out that the data sample is too short for claiming 80% success rate, you started new argument about the relevance of previous data. I did not say 10 years, I said not enough sample size which will be pointed out by anyone who has given your paper a serious glance.
This is the reason I said this is more of a philosophical paper.

The current price of bitcoin is determined by the collective actions of all the traders’ participating, whether they buy or sell is what causes the price of bitcoin to fluctuate, no amount of historical data, whether from years ago and days ago determines the current price of bitcoin, and definitely the superposition principle does not determine the price of bitcoin, don’t know where you got that impression. As to your opinion about four weeks of data being too small of a dataset to claim 80% accuracy, we particularly used a smaller dataset to forecast the following week because for finding the trend of the short forecast horizon the more recent historical data gives more insight as opposed to data from a year or two ago. Essentially, for statistical analysis, indeed using large amounts of data might be of better use, but for forecasting it’s simply not just the bigger the better. And lastly our methodology can be utilized to time series forecasting generically, stocks and bitcoin are just one case study, our algorithm can forecast total sales data; retail demand, QSAR of drug design, scientific discovery, etc…, basically any form of time series forecast data. Our algorithm has been tested on many real-world datasets, so we don’t get your point of why you keep persistently mention that our paper is a philosophical one.