Support vector fuzzy parallel embedded system
There are many problems faced by fund managers in managing a portfolio. The common problems consist of not knowing how to allocate assets, which stocks to include, and how to rebalance assets in the portfolio. Most portfolios today are managed by active fund managers. The issue with active portfolio...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project (FYP) |
| Language: | English |
| Published: |
Nanyang Technological University
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/156492 |
| _version_ | 1826120119643799552 |
|---|---|
| author | Book, Jeremy Kay Yip |
| author2 | Quek Hiok Chai |
| author_facet | Quek Hiok Chai Book, Jeremy Kay Yip |
| author_sort | Book, Jeremy Kay Yip |
| collection | NTU |
| description | There are many problems faced by fund managers in managing a portfolio. The common problems consist of not knowing how to allocate assets, which stocks to include, and how to rebalance assets in the portfolio. Most portfolios today are managed by active fund managers. The issue with active portfolio management by an active fund manager is often plagued by limitations and shortcomings, such as limited processing capabilities of the human brain and the presence of cognitive biases such as overconfidence that can be developed over time due to previous successes.
Artificial intelligence (AI) and Machine learning (ML) have been adopted by fund managers to assist with their active portfolio management process [1]. The predictive ability of AI and ML can provide fund managers with forecasted information in the stock market, allowing them to make early informed decisions for upside potential profits.
However, AI and ML lack interpretability regarding how their outputs are derived and thus function as black boxes [3]. The black box nature of AI and ML makes it seem unreliable and uncertain. Without a proper explanation of the predicted output, humans tend to feel sceptical and doubtful. Hence it is desirable to have an architecture that has predictive ability and provides interpretations.
This paper proposes and illustrates an architecture, Support Vector Fuzzy Parallel Embedded System (SVFPS) by incorporating a fuzzy system embedded with machine learning. The proposed architecture functions as a predictive model with an ability to form highly intuitive IF-THEN fuzzy rules to provide linguistic insights of how outputs are derived.
The effectiveness of the proposed architecture, Support Vector Fuzzy Parallel Embedded System (SVFPS) is evaluated by incorporating SVFPS into a portfolio management system with several sector Exchange-Traded Funds (ETFs). The experimental results showed that the portfolio management incorporated with the proposed SVFPS has outperformed benchmarks of commonly used investing strategies. |
| first_indexed | 2024-10-01T05:10:59Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/156492 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T05:10:59Z |
| publishDate | 2022 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1564922022-04-17T13:13:14Z Support vector fuzzy parallel embedded system Book, Jeremy Kay Yip Quek Hiok Chai School of Computer Science and Engineering ASHCQUEK@ntu.edu.sg Engineering::Computer science and engineering There are many problems faced by fund managers in managing a portfolio. The common problems consist of not knowing how to allocate assets, which stocks to include, and how to rebalance assets in the portfolio. Most portfolios today are managed by active fund managers. The issue with active portfolio management by an active fund manager is often plagued by limitations and shortcomings, such as limited processing capabilities of the human brain and the presence of cognitive biases such as overconfidence that can be developed over time due to previous successes. Artificial intelligence (AI) and Machine learning (ML) have been adopted by fund managers to assist with their active portfolio management process [1]. The predictive ability of AI and ML can provide fund managers with forecasted information in the stock market, allowing them to make early informed decisions for upside potential profits. However, AI and ML lack interpretability regarding how their outputs are derived and thus function as black boxes [3]. The black box nature of AI and ML makes it seem unreliable and uncertain. Without a proper explanation of the predicted output, humans tend to feel sceptical and doubtful. Hence it is desirable to have an architecture that has predictive ability and provides interpretations. This paper proposes and illustrates an architecture, Support Vector Fuzzy Parallel Embedded System (SVFPS) by incorporating a fuzzy system embedded with machine learning. The proposed architecture functions as a predictive model with an ability to form highly intuitive IF-THEN fuzzy rules to provide linguistic insights of how outputs are derived. The effectiveness of the proposed architecture, Support Vector Fuzzy Parallel Embedded System (SVFPS) is evaluated by incorporating SVFPS into a portfolio management system with several sector Exchange-Traded Funds (ETFs). The experimental results showed that the portfolio management incorporated with the proposed SVFPS has outperformed benchmarks of commonly used investing strategies. Bachelor of Engineering (Computer Science) 2022-04-17T13:13:13Z 2022-04-17T13:13:13Z 2022 Final Year Project (FYP) Book, J. K. Y. (2022). Support vector fuzzy parallel embedded system. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/156492 https://hdl.handle.net/10356/156492 en SCSE21-0431 application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Computer science and engineering Book, Jeremy Kay Yip Support vector fuzzy parallel embedded system |
| title | Support vector fuzzy parallel embedded system |
| title_full | Support vector fuzzy parallel embedded system |
| title_fullStr | Support vector fuzzy parallel embedded system |
| title_full_unstemmed | Support vector fuzzy parallel embedded system |
| title_short | Support vector fuzzy parallel embedded system |
| title_sort | support vector fuzzy parallel embedded system |
| topic | Engineering::Computer science and engineering |
| url | https://hdl.handle.net/10356/156492 |
| work_keys_str_mv | AT bookjeremykayyip supportvectorfuzzyparallelembeddedsystem |