PREreview of Do Technical Indicators Improve Deep Learning Forecasts? An Empirical Ablation Study Across Asset Classes
- Published
- DOI
- 10.5281/zenodo.19521824
- License
- CC BY 4.0
Biggest concern is whether any form of leakage could occur when constructing indicators. There is high possibilty that leakage may occur in such experiments. And even small leakage can change the result of paper significantly.
Many of the technical indicators are not fixed. They depends heavily on lookback period. 50 day SMA, 200 day SMA and 100 day SMA completely moves differently. So if the paper is using any particular lookback window and ignoring others, it should be mentioned on the paper.
Thing I find most interesting is volatility only helps in Forex. I would love to read more about it becuase it is not clearly explained in the abstract.
Competing interests
The author declares that they have no competing interests.
Use of Artificial Intelligence (AI)
The author declares that they did not use generative AI to come up with new ideas for their review.