Scribd
Upload
131 views8 pages

Information Management: End Term Project

The document describes two proposed models - a Trading Model and an Investment Model - to automate reactions to news and make stock investment suggestions. The Trading Model would use text mining on news feeds and social media to react as fast as possible to news and make buy, sell, or short sell recommendations through a trading system. The Investment Model would use text mining on financial web information and analysts' reports to suggest buy, sell, or hold decisions. Both models would be tested using reinforced learning - the Trading Model aims for 70% accuracy and the Investment Model aims for 80% accuracy over a two year period.

Uploaded by

abhinavdeb
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
131 views8 pages

Information Management: End Term Project

The document describes two proposed models - a Trading Model and an Investment Model - to automate reactions to news and make stock investment suggestions. The Trading Model would use text mining on news feeds and social media to react as fast as possible to news and make buy, sell, or short sell recommendations through a trading system. The Investment Model would use text mining on financial web information and analysts' reports to suggest buy, sell, or hold decisions. Both models would be tested using reinforced learning - the Trading Model aims for 70% accuracy and the Investment Model aims for 80% accuracy over a two year period.

Uploaded by

abhinavdeb
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

Information Management

End Term Project


Abhinav Deb (170301002)
Arya Kishore Sarbadhikary (170301015)
Nilesh Dutta (170301030)
Soumya Basu (170301037)
Subasish Ansuman Sharma (170301038)
WHY
1. Share Market react so fast to news, that it doesn’t give any time to
human beings to react

2. It is impossible for a human being to react to each and every news in


real time.

3. Analytics can be a great tool for equity investment


What we are solving
1. Automate the process of reacting to the news as fast as possible.

2. Create a model to keep a track of all the news items and react
accordingly.

3. Create a model to use analytics and AI to suggest stock investments.


Solution
Trading Model
Stock Exchange
News Feed
Regular Stock
Performance
Text mining Trading System Interpreter
Social Media

Buy/Sell/Short Sell Event Information

Global
Algorithm
Market
Solution
Investment Model

World Wide
Web
Buy/Sell/Hold
Financial
Web Investing System Stock Exchange
Information Text mining Algorithm
Archive

Analysts Portfolio Information


Report
Interpreter
Testing
Trading Model Testing:
• The Trading model will be tested against the virtual trading platform and the
algorithm will be self modified with reinforced learning.
• 70% accuracy will be considered as the final target accuracy.

Investment Model Testing:


• The algorithm will be tested with each quarter performance information of
the stock in the portfolio, fed back to the algorithm for reinforced learning.
• Accuracy target is 80% for during a period of 2 years.
Context to Solution Mapping
Context: Share Market react so fast to news, that it doesn’t give any
time to human beings to react
Solution: Trading Model
Context: It is impossible for a human being to react to each and every
news in real time.
Solution: Trading Model
Context: Analytics can be a great tool for equity investment
Solution: Investment Model
Thank You!

You might also like