Using Hidden Markov Model to Predict Human Actions with Swarm Intelligence

Zhicheng Lu, Yuk Ying Chung, Henry Wing Fung Yeung, Seid Miad Zandavi, Weiming Zhi, Wei Chang Yeh

Research output: Book chapter/Published conference paperConference paperpeer-review

4 Citations (Scopus)

Abstract

This paper proposed a novel algorithm which named Randomized Particle Swarm Optimization (RPSO) to optimize HMM for human activity prediction. The experiments designed in this paper are the classification of human activity using two data sets. The first testing data is from the TUM Kitchen Data Set and the other is the Human Activity Recognition using the Smartphone Data Set from UCI Machine Learning Repository. Based on the comparison of the accuracies for the conventional HMM and optimized HMM, a conclusion can be drawn that the proposed RPSO can help HMM to achieve higher accuracy for human action recognition. Our results show that RPSO-HMM can improve 15% accuracy in human activity recognition and prediction when compared to the traditional HMM.

Original languageEnglish
Title of host publicationNeural Information Processing - 24th International Conference, ICONIP 2017, Proceedings
EditorsDerong Liu, Shengli Xie, Yuanqing Li, El-Sayed M. El-Alfy, Dongbin Zhao
PublisherSpringer-Verlag Italia Srl
Pages21-30
Number of pages10
ISBN (Print)9783319700922
DOIs
Publication statusPublished - 2017
Event24th International Conference on Neural Information Processing, ICONIP 2017 - Guangzhou, China
Duration: 14 Nov 201718 Nov 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10637 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference24th International Conference on Neural Information Processing, ICONIP 2017
Country/TerritoryChina
CityGuangzhou
Period14/11/1718/11/17

Fingerprint

Dive into the research topics of 'Using Hidden Markov Model to Predict Human Actions with Swarm Intelligence'. Together they form a unique fingerprint.

Cite this