Stock Trading Accelerator using Reinforcement Learning

Aishwarya Srinivasan
2 min readOct 4, 2021

TRADING! 🤑🤑

A use-case that gets more complicated than we can ever imagine. There are so many data signals that could impact the market movement. It isn’t just a simple time-series use-case, there are so many factors that could trigger the market growth or fall. In recent times, we have seen community sentiments affecting the market valuation of companies. Similarly, there could be factors like company mergers, acquisitions, partnerships, new technology launches, and many more factors which could be insightful factors to evaluate the market movement.

As fancy as the trading algorithms could get, there is always a fact of models getting outdated and requires intensive research for them to be modified. One of the solutions is Reinforcement Learning models, which can over time, learn new patterns, and update themselves to always look for these trends.

The factors that once played a crucial role in the stock market movement might now not be the same and there could be more input features that could be more relevant. With Reinforcement Learning models you can create such adaptive and robust models which work on a set of parameters and predict the next best action for you to take in the market position.

We have built a Stock Trader Accelerator which can help you start your journey with stock trading using Reinforcement Learning.

Check out the accelerator here: https://community.ibm.com/accelerators/catalog/content/Stock-Trading

--

--

Aishwarya Srinivasan

LinkedIn Top Voice 2020- Data Science || MS Data Science - Columbia University || IBM- Data Science Elite || Unicorn in Data Science || Scikit-Learn Contributor