Smart home device usage prediction using pattern matching and reinforcement learning

Mamun Bin Ibne Reaz; A. Assim; M. I. Ibrahimy; Florence Chiao Mei Choong; Faisal Mohd-Yasin

Smart home device usage prediction using pattern matching and reinforcement learning

Mamun Bin Ibne Reaz, A. Assim, M. I. Ibrahimy, Florence Chiao Mei Choong, Faisal Mohd-Yasin

School of Engineering & Physical Sciences

Research output: Contribution to conference › Paper › peer-review

Abstract

Future Smart-Home device usage prediction is a very important module in artificial intelligence. The technique involves analyzing the user performed actions history and applies mathematical methods to predict the most feasible next user action. Present a new algorithm of user action prediction based on pattern matching and techniques of reinforcement learning. Synthetic data had been used to test the algorithm and the result shows that the algorithm presented in this paper performs 20% better than the current available techniques.

Original language	English
Pages	1092-1098
Publication status	Published - 2008
Event	7th International Conference on System Identification and Control Problems - Moscow, Russian Federation Duration: 28 Jan 2008 → 31 Jan 2008 http://www.sicpro.org/sicpro08/code/e08_01.htm#PROC

Conference

Conference	7th International Conference on System Identification and Control Problems
Abbreviated title	(SCIPRO’08)
Country/Territory	Russian Federation
City	Moscow
Period	28/01/08 → 31/01/08
Internet address	http://www.sicpro.org/sicpro08/code/e08_01.htm#PROC

Keywords

smart-home
reinforcement-learning
multi-agent
pattern-matching

Cite this

@conference{040e0c4fee674dd399d7342ec69ad7aa,

title = "Smart home device usage prediction using pattern matching and reinforcement learning",

abstract = "Future Smart-Home device usage prediction is a very important module in artificial intelligence. The technique involves analyzing the user performed actions history and applies mathematical methods to predict the most feasible next user action. Present a new algorithm of user action prediction based on pattern matching and techniques of reinforcement learning. Synthetic data had been used to test the algorithm and the result shows that the algorithm presented in this paper performs 20% better than the current available techniques.",

keywords = "smart-home, reinforcement-learning, multi-agent, pattern-matching",

author = "Reaz, {Mamun Bin Ibne} and A. Assim and Ibrahimy, {M. I.} and Choong, {Florence Chiao Mei} and Faisal Mohd-Yasin",

year = "2008",

language = "English",

pages = "1092--1098",

note = "7th International Conference on System Identification and Control Problems , (SCIPRO{\textquoteright}08) ; Conference date: 28-01-2008 Through 31-01-2008",

url = "http://www.sicpro.org/sicpro08/code/e08_01.htm#PROC",

}

TY - CONF

T1 - Smart home device usage prediction using pattern matching and reinforcement learning

AU - Reaz, Mamun Bin Ibne

AU - Assim, A.

AU - Ibrahimy, M. I.

AU - Choong, Florence Chiao Mei

AU - Mohd-Yasin, Faisal

PY - 2008

Y1 - 2008

N2 - Future Smart-Home device usage prediction is a very important module in artificial intelligence. The technique involves analyzing the user performed actions history and applies mathematical methods to predict the most feasible next user action. Present a new algorithm of user action prediction based on pattern matching and techniques of reinforcement learning. Synthetic data had been used to test the algorithm and the result shows that the algorithm presented in this paper performs 20% better than the current available techniques.

AB - Future Smart-Home device usage prediction is a very important module in artificial intelligence. The technique involves analyzing the user performed actions history and applies mathematical methods to predict the most feasible next user action. Present a new algorithm of user action prediction based on pattern matching and techniques of reinforcement learning. Synthetic data had been used to test the algorithm and the result shows that the algorithm presented in this paper performs 20% better than the current available techniques.

KW - smart-home

KW - reinforcement-learning

KW - multi-agent

KW - pattern-matching

M3 - Paper

SP - 1092

EP - 1098

T2 - 7th International Conference on System Identification and Control Problems

Y2 - 28 January 2008 through 31 January 2008

ER -

Smart home device usage prediction using pattern matching and reinforcement learning

Abstract

Conference

Keywords

Fingerprint

Cite this