Been half a month learning about quant in finance and came through your channel. Its really awesome. Why don't you guys post regularly? Do you also have a learning program by any chance?
I can see how this might be used to make ML projections, but can such a method also be used for live-online-streaming time series without leaking information from the future?
Hi Michael, this is a question! I had the same question when I was first learning it as well! Fractional Differencing helps to make the data stationary while preserving as much information in the data as possible. If you are using a trend following technical indicator such as MACD, Moving Averages, etc.., it would be best to apply them after fractionally differencing your data, in order to get the most clear trend possible.
Any chance you could do a write up and video on using wavelet transformations for denoising and the complexities of eliminating or detecting edge effects? There’s a dearth of literature surrounding this.
Bumping your channel! This is great work
Thank you, that is interesting and nicely presented. Looking forward to more content
Thank you for watching! Let us know if you have anything you would be interested in seeing!
Common guys !! It has been a year we want new stuff ❤️❤️
Been half a month learning about quant in finance and came through your channel. Its really awesome.
Why don't you guys post regularly? Do you also have a learning program by any chance?
Would love to see a video on meta labeling!
I can see how this might be used to make ML projections, but can such a method also be used for live-online-streaming time series without leaking information from the future?
Should keep making vids
Thank you! Do you have any topics you would be interested in us covering?
I liked your video. I've always wondered, do you create technical indicators after fractionally differencing or before?
Hi Michael, this is a question! I had the same question when I was first learning it as well! Fractional Differencing helps to make the data stationary while preserving as much information in the data as possible. If you are using a trend following technical indicator such as MACD, Moving Averages, etc.., it would be best to apply them after fractionally differencing your data, in order to get the most clear trend possible.
Makes sense. Thanks for the reply!
Any chance you could do a write up and video on using wavelet transformations for denoising and the complexities of eliminating or detecting edge effects? There’s a dearth of literature surrounding this.
this shit is underrated
Thank you! Let us know if you have anything you'd like us to cover