1
$\begingroup$

Currently, what are the most popular and effective approaches to leveraging AI for stock price prediction?

It seems like there could be several approaches and problem formulations:

  • Supervised learning:
  • Regression: predict the stock price directly
  • Classification: predict whether the stock price goes up or down
  • Unsupervised learning: find clusters of stocks that move together
  • Reinforcement learning: let the agent directly maximize its stock market return
  • Other AI methods: rules, symbolic systems, etc.

Which are most popular/performant? Are there other ways that people are using machine learning in stock trading (sentiment analysis on financial statements, news, etc.)?

$\endgroup$

4 Answers 4

1
$\begingroup$

Your list is complete for what is considered 'popular' by most practitioners who apply AI for stock trading. Supervised learning and rule learning are at the top for accuracy. There are more academic papers published on classifiers than on regression approaches; classifiers are typically more accurate than regressors.

$\endgroup$
1
$\begingroup$

NLP is one of the tools you may use for stock prediction. Here, are a couple of articles to help you get started:

$\endgroup$
1
  • $\begingroup$ Do you know any company that has published the way they are using NLP to predict the price of a stock (in the short team I guess)? $\endgroup$ Commented May 25, 2021 at 7:56
1
$\begingroup$

Usecases I came across:

  1. As mentioned by Saurav, the use of NLP is definitely a use case. Adding to his source, you can check out the Two Sigma's Kaggle competition. Going through their careers page it is evident that use of NLP is prominent. Cofounder of Kavout, Alex Lu reconfirms that in https://emerj.com/ai-podcast-interviews/artificial-intelligence-in-stock-trading-future-trends-and-applications/

Two Sigma Challenge Kaggle: https://www.kaggle.com/c/two-sigma-financial-news

  1. ML to determine whether some large market player is rebalancing or liquidating his or her portfolio. Source: https://qr.ae/pGnoGx
  2. I have heard about ML being used in Statistical Arbitrage(Pairs trading) also. Example case: http://cs229.stanford.edu/proj2019spr/report/26.pdf
  3. Further, I am aware of use of ML for hedging purposes in a couple of banks. It's called deep hedging. Source: https://www.maths.ox.ac.uk/system/files/attachments/2019%2004%2024%20Deep%20Hedging%20Frontiers%20Imperial%202.1.pdf
$\endgroup$
1
$\begingroup$

One technique I've used in the past (I don't know how mainstream this is) is to use unsupervised learning (self-organizing maps) to identify commonly occuring features in stock charts, and then use those features as input to a traditional back propagation network. I found I was able to get equivalent results with slightly smaller network size having done this, but it wasn't a huge gain.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .