Financial prediction using fuzzy inference system
Financial Markets have been increasingly attractive as the ways of investing in stocks, commodities and such have become much easier. The stories and advertisements of successful investments and fast ways to get rich has attracted more and more investors. With the advancement of technologies and bre...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/66731 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Financial Markets have been increasingly attractive as the ways of investing in stocks, commodities and such have become much easier. The stories and advertisements of successful investments and fast ways to get rich has attracted more and more investors. With the advancement of technologies and breakthrough in the field of Machine Learning and predictive algorithms, there might be a tunnel of light towards a more rigid prediction in the world of Financial Markets.
The project aimed to produce a financial prediction system which makes investment decision for profiteering. As Type-1 fuzzy systems are not effective in handling uncertainties in data, the project uses a Simplified Interval Type-2 Fuzzy Neural Networks algorithm which is better at handling uncertainties.
On the track of Technical Analysis used by most of the investors for predicting future prices based on Historical prices of the company, the project makes use of a few financial indicators to predict and decide the trends of the market in the near future. It does so by looking through a small window of 5 – 10 days, predicting the trend line and find the lowest and highest points for buying and selling of the stock.
With a functional implementation, the trend line produced by the prediction variable was although not performing at the maximize level, but it still provided an overall justifiable result in getting a good return of investments. The issues identified for the project is probably the way the prediction variable is calculated and the lack of automated mechanism to adjust thresholds.
Some recommendation to tackle the problem of the current implementation could include drawing data from other sources like social media platform, global news, company news, etc. to judge the company in the sentimental analysis aspect. |
|---|