Predicting S&P 500 based on its constituents and their social media derived sentiment
- Submitting institution
-
Goldsmiths' College
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 3377
- Type
- E - Conference contribution
- DOI
-
10.1007/978-3-030-28377-3_12
- Title of conference / published proceedings
- Computational Collective Intelligence: 11th International Conference, ICCCI 2019, Hendaye, France, September 4–6, 2019, Proceedings, Part I
- First page
- 142
- Volume
- 11683
- Issue
- -
- ISSN
- 0302-9743
- Open access status
- Other exception
- Month of publication
- August
- Year of publication
- 2019
- URL
-
http://research.gold.ac.uk/id/eprint/26368/
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
5
- Research group(s)
-
-
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This work proposes, for the first time, an approach to predicting S&P 500 based on the closing stock prices and sentiment data of the S&P 500 constituents. We introduce an unprecedented approach to predicting S&P500, as a stock market and sentiment-analysis application. The research benefits PhD training in Data Science and inspires the candidate's banking profession. This work proposes an original, state-of-the-art predictive modelling approach based on Jordan and Elman recurrent neural-networks, and implements an efficient hybrid trading system that is a viable solution for maximising investment portfolios. The paper is published in the Springer proceedings of ICCCI.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -