Topic: Data Analysts Regression Model to 1 Million USD per Bitcoin (Read 99 times)
First of all, take the analyst seriously, no they’re in the end just analyst, they’re not trustworthy, when thing turn south they don’t have anything to back their blatant lies, I hope this is not another McAfee recommended analyst.
I am using a linear regression model with function (as of today) "f(x) = 10^(x*0011641)
Is this an exponential regression? And the predictor is only the number of days? C'mon man!Yes. The beauty is it`s simplicity and high effectiveness and high accuracy. The root mean squared of the given equation is considered very high. I could use more complex models and even machine learning or so called black box methods with neural networks that might give us some short term predictions but not in any way are they more accurate in long term predictions. If you take for example Moore`s law- you would have been able to predict todays computing power with very good precision in 1975. Just like computer chips, Bitcoin is a technology and it`s real value (not price) increases at a constant pace. Regardless of that, I will still update this thread with more models and will give out some trading strategies with backtested logs. The purpose of this thread is to help folks make more educated decisions and reduce risk. I highly recommend subscribing, there is a lot more to come.
Did you already made your software? I've heard of that idea before about forecasting bitcoin's price through a software and now here you are, making that possible.
I would appreciate if someone could tell me how to embed these .png-s to the post properly.
You need to be at least jr.member or avail the copper membership for you be able to post images. But if someone who's on that rank quotes the link, the images will appear just like this. I'll quote one for you.Every single plot in this thread is produced by the software I wrote. Nothing is calculated manually, everything is computed. Actually this is just the start of the project. There is a lot more to come.
Thanks for the info. I have now upgraded my account to copper.
I am using a linear regression model with function (as of today) "f(x) = 10^(x*0011641)
Is this an exponential regression? And the predictor is only the number of days? C'mon man!It doesn't seem valid for me, and a simple program like ms excel would do the job. However, if it's a plugin for trading, it might be useful for some traders.
Did you already made your software? I've heard of that idea before about forecasting bitcoin's price through a software and now here you are, making that possible.
I would appreciate if someone could tell me how to embed these .png-s to the post properly.
You need to be at least jr.member or avail the copper membership for you be able to post images. But if someone who's on that rank quotes the link, the images will appear just like this. I'll quote one for you.
Hi Bitcoin community!
I represent here myself as "PizzaRoma" and I am a business analyst by trade. Last couple of months I used my downtime to write a software that analysis and forecasts Bitcoin price. I am using a linear regression model with function (as of today) "f(x) = 10^(x*0011641) where f(x) is today`s expected price and "x" is number of days since "2010-07-17". Root Mean Squared of the model is 0.8752 which is considered very good. Today we are just -12.8593% behind the trend. With current trend the forecasted date for 1M USD/BTC is "2024-08-26".
I represent here myself as "PizzaRoma" and I am a business analyst by trade. Last couple of months I used my downtime to write a software that analysis and forecasts Bitcoin price. I am using a linear regression model with function (as of today) "f(x) = 10^(x*0011641) where f(x) is today`s expected price and "x" is number of days since "2010-07-17". Root Mean Squared of the model is 0.8752 which is considered very good. Today we are just -12.8593% behind the trend. With current trend the forecasted date for 1M USD/BTC is "2024-08-26".
Jump to: