Summary
International Symposium on Nonlinear Theory and Its Applications
2016
Session Number:A3L-B
Session:
Number:A3L-B-3
Biased Reactions to Abnormal Stock Prices Detected by Autoencoder During Daytime and Nighttime
Hiroyuki Gotou, Tomoya Suzuki,
pp.-
Publication Date:2016/11/27
Online ISSN:2188-5079
Summary:
We applied an autoencoder to learn the relationships among all stocks, and considered the stock prices that the autoencoder cannot restore as abnormal prices. Then, we identified that the price movements following the abnormal prices caused during the nighttime are clearly biased. As its application, we proposed contrarian trading strategies and confirmed each validity by some statistical significance tests.