CMOS compatible building blocks for reservoir computing

This project was done to learn more about what a PRC is and to compare it against the traditional RNN models. Chaos maps are chosen to be the input due to their chaotic behaviour which serves as a great benchmark for our models. They also serve to test the model’s robustness to noise as well as thei...

Full description

Saved in:
Bibliographic Details
Main Author: Wee, Jun Lin
Other Authors: Ang Diing Shenp
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2024
Subjects:
Online Access:https://hdl.handle.net/10356/176881
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-176881
record_format dspace
spelling sg-ntu-dr.10356-1768812024-05-31T15:45:21Z CMOS compatible building blocks for reservoir computing Wee, Jun Lin Ang Diing Shenp School of Electrical and Electronic Engineering EDSAng@ntu.edu.sg Engineering This project was done to learn more about what a PRC is and to compare it against the traditional RNN models. Chaos maps are chosen to be the input due to their chaotic behaviour which serves as a great benchmark for our models. They also serve to test the model’s robustness to noise as well as their capability to process temporal and sequential data. In this project, Duffing Map, Gingerbreadman Map and Hénon Map are used as the testbench. The PRC model and RNN model are pitted against each other in their capability to predict different chaos maps. By measuring the NRMSE value of the output respectively, this experiment proves to show whether a PRC model is much more efficient than a RNN model in using their computational resources. Bachelor's degree 2024-05-31T00:02:41Z 2024-05-31T00:02:41Z 2024 Final Year Project (FYP) Wee, J. L. (2024). CMOS compatible building blocks for reservoir computing. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/176881 https://hdl.handle.net/10356/176881 en application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering
spellingShingle Engineering
Wee, Jun Lin
CMOS compatible building blocks for reservoir computing
description This project was done to learn more about what a PRC is and to compare it against the traditional RNN models. Chaos maps are chosen to be the input due to their chaotic behaviour which serves as a great benchmark for our models. They also serve to test the model’s robustness to noise as well as their capability to process temporal and sequential data. In this project, Duffing Map, Gingerbreadman Map and Hénon Map are used as the testbench. The PRC model and RNN model are pitted against each other in their capability to predict different chaos maps. By measuring the NRMSE value of the output respectively, this experiment proves to show whether a PRC model is much more efficient than a RNN model in using their computational resources.
author2 Ang Diing Shenp
author_facet Ang Diing Shenp
Wee, Jun Lin
format Final Year Project
author Wee, Jun Lin
author_sort Wee, Jun Lin
title CMOS compatible building blocks for reservoir computing
title_short CMOS compatible building blocks for reservoir computing
title_full CMOS compatible building blocks for reservoir computing
title_fullStr CMOS compatible building blocks for reservoir computing
title_full_unstemmed CMOS compatible building blocks for reservoir computing
title_sort cmos compatible building blocks for reservoir computing
publisher Nanyang Technological University
publishDate 2024
url https://hdl.handle.net/10356/176881
_version_ 1814047438454915072