A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction
We propose a new nature- and neuro-science-inspired algorithm for spatiotemporal learning and prediction based on sequential recall and vector symbolic architecture. A key novelty is the learning of spatial and temporal patterns as decoupled concepts where the temporal pattern sequences are construc...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Biomimetics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-7673/9/3/175 |
| _version_ | 1797241914813054976 |
|---|---|
| author | Thimal Kempitiya Damminda Alahakoon Evgeny Osipov Sachin Kahawala Daswin De Silva |
| author_facet | Thimal Kempitiya Damminda Alahakoon Evgeny Osipov Sachin Kahawala Daswin De Silva |
| author_sort | Thimal Kempitiya |
| collection | DOAJ |
| description | We propose a new nature- and neuro-science-inspired algorithm for spatiotemporal learning and prediction based on sequential recall and vector symbolic architecture. A key novelty is the learning of spatial and temporal patterns as decoupled concepts where the temporal pattern sequences are constructed using the learned spatial patterns as an alphabet of elements. The decoupling, motivated by cognitive neuroscience research, provides the flexibility for fast and adaptive learning with dynamic changes to data and concept drift and as such is better suited for real-time learning and prediction. The algorithm further addresses several key computational requirements for predicting the next occurrences based on real-life spatiotemporal data, which have been found to be challenging with current state-of-the-art algorithms. Firstly, spatial and temporal patterns are detected using unsupervised learning from unlabeled data streams in changing environments; secondly, vector symbolic architecture (VSA) is used to manage variable-length sequences; and thirdly, hyper dimensional (HD) computing-based associative memory is used to facilitate the continuous prediction of the next occurrences in sequential patterns. The algorithm has been empirically evaluated using two benchmark and three time-series datasets to demonstrate its advantages compared to the state-of-the-art in spatiotemporal unsupervised sequence learning where the proposed ST-SOM algorithm is able to achieve 45% error reduction compared to HTM algorithm. |
| first_indexed | 2024-04-24T18:30:54Z |
| format | Article |
| id | doaj.art-1f5213a89db74937b96cd56e9e308b18 |
| institution | Directory Open Access Journal |
| issn | 2313-7673 |
| language | English |
| last_indexed | 2024-04-24T18:30:54Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Biomimetics |
| spelling | doaj.art-1f5213a89db74937b96cd56e9e308b182024-03-27T13:27:42ZengMDPI AGBiomimetics2313-76732024-03-019317510.3390/biomimetics9030175A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and PredictionThimal Kempitiya0Damminda Alahakoon1Evgeny Osipov2Sachin Kahawala3Daswin De Silva4Centre for Data Analytics and Cognition, La Trobe University, Melbourne, VIC 3086, AustraliaCentre for Data Analytics and Cognition, La Trobe University, Melbourne, VIC 3086, AustraliaDepartment of Computer Science, Electrical and Space Engineering, Luleå University of Technology, 971 87 Luleå, SwedenCentre for Data Analytics and Cognition, La Trobe University, Melbourne, VIC 3086, AustraliaCentre for Data Analytics and Cognition, La Trobe University, Melbourne, VIC 3086, AustraliaWe propose a new nature- and neuro-science-inspired algorithm for spatiotemporal learning and prediction based on sequential recall and vector symbolic architecture. A key novelty is the learning of spatial and temporal patterns as decoupled concepts where the temporal pattern sequences are constructed using the learned spatial patterns as an alphabet of elements. The decoupling, motivated by cognitive neuroscience research, provides the flexibility for fast and adaptive learning with dynamic changes to data and concept drift and as such is better suited for real-time learning and prediction. The algorithm further addresses several key computational requirements for predicting the next occurrences based on real-life spatiotemporal data, which have been found to be challenging with current state-of-the-art algorithms. Firstly, spatial and temporal patterns are detected using unsupervised learning from unlabeled data streams in changing environments; secondly, vector symbolic architecture (VSA) is used to manage variable-length sequences; and thirdly, hyper dimensional (HD) computing-based associative memory is used to facilitate the continuous prediction of the next occurrences in sequential patterns. The algorithm has been empirically evaluated using two benchmark and three time-series datasets to demonstrate its advantages compared to the state-of-the-art in spatiotemporal unsupervised sequence learning where the proposed ST-SOM algorithm is able to achieve 45% error reduction compared to HTM algorithm.https://www.mdpi.com/2313-7673/9/3/175self-organizing mapsspatiotemporal sequence learningvector symbolic architectureshierarchical temporal memory |
| spellingShingle | Thimal Kempitiya Damminda Alahakoon Evgeny Osipov Sachin Kahawala Daswin De Silva A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction Biomimetics self-organizing maps spatiotemporal sequence learning vector symbolic architectures hierarchical temporal memory |
| title | A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction |
| title_full | A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction |
| title_fullStr | A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction |
| title_full_unstemmed | A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction |
| title_short | A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction |
| title_sort | two layer self organizing map with vector symbolic architecture for spatiotemporal sequence learning and prediction |
| topic | self-organizing maps spatiotemporal sequence learning vector symbolic architectures hierarchical temporal memory |
| url | https://www.mdpi.com/2313-7673/9/3/175 |
| work_keys_str_mv | AT thimalkempitiya atwolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT dammindaalahakoon atwolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT evgenyosipov atwolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT sachinkahawala atwolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT daswindesilva atwolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT thimalkempitiya twolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT dammindaalahakoon twolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT evgenyosipov twolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT sachinkahawala twolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction AT daswindesilva twolayerselforganizingmapwithvectorsymbolicarchitectureforspatiotemporalsequencelearningandprediction |