Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/156568
Title: Fuzzy C-means long short-term memory (FCMLSTM) with application in exchange-traded funds (ETFs)
Authors: Siek, Ming Kang
Keywords: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
Issue Date: 2022
Publisher: Nanyang Technological University
Source: Siek, M. K. (2022). Fuzzy C-means long short-term memory (FCMLSTM) with application in exchange-traded funds (ETFs). Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/156568
Abstract: Breakthrough in computational power and together with the abundance of large datasets available had contributed a significant role in stimulating the technological advancement in artificial neural networks. While artificial neural networks may be a powerful tool, it is also criticized as a black box. The models are difficult for humans to interpret directly as it creates complex mathematical functions within itself. To address this issue, many have investigated into fuzzy neural networks which combines the interpretability of fuzzy system and performance of neural networks. This paper proposes a system using Fuzzy C-Means (FCM) algorithm and embedded with a deep Long Short-Term Memory (LSTM) neural network to form a FCMLSTM system that handles time series problems. Our system uses FCM to form IF-THEN rules which provide interpretability to results. Fuzzy input generated by FCM is forwarded to the LSTM model to generate a fuzzy output. The fuzzy output will be interpreted using the Mamdani Inference Model and IF-THEN rules to generate a crisp output using the centre of gravity defuzzification method. Our FCMLSTM system’s performance is evaluated using metrics to measure the accuracy and the trend of the outputs. The performance of our system in predicting trend showed promising results and led to the investigation of utilising our system to incorporate together with trend indicators to form our trading strategy predicted Moving Average Divergence Histogram (MACDH). Our trading strategy will trade based on the buy and sell signals and is applied to conduct a portfolio rebalancing strategy. The results will be compared to the buy & hold strategy, hindsight MACDH and MACDH evaluated by the return of interest and annualised return of interest.
URI: https://hdl.handle.net/10356/156568
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
Amended FYP Report 150422.pdf
  Restricted Access
2.32 MBAdobe PDFView/Open

Page view(s)

9
Updated on May 19, 2022

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.