Partly Locality Sensitive Hashing を用いた時系列データからの高頻度パターン抽出

Translated title of the contribution: Detecting Frequent Patterns in Time Series Data using Partly Locality Sensitive Hashing

小川原 光一, 田邉 康史, 倉爪 亮, 長谷川 勉

Research output: Contribution to journalArticle

Abstract

Frequent patterns in time series data are useful clues to learn previously unknown events in an unsupervised way. In this paper, we propose a method for detecting frequent patterns in long time series data efficiently. The major contribution of the paper is two-fold: (1) Partly Locality Sensitive Hashing (PLSH) is proposed to find frequent patterns efficiently and (2) the problem of finding consecutive time frames that have a large number of frequent patterns is formulated as a combinatorial optimization problem which is solved via Dynamic Programming (DP) in polynomial time <i>O</i> (<i>N</i> <sup>1+1/α</sup>) thanks to PLSH where <i>N</i> is the total amount of data. The proposed method was evaluated by detecting frequent whole body motions in a video sequence as well as by detecting frequent everyday manipulation tasks in motion capture data.
Original languageJapanese
Pages (from-to)67-76
Number of pages10
Journal日本ロボット学会誌
Volume29
Issue number1
DOIs
Publication statusPublished - Jan 15 2011

Fingerprint

Time series
Combinatorial optimization
Dynamic programming
Data acquisition
Polynomials

Cite this

Partly Locality Sensitive Hashing を用いた時系列データからの高頻度パターン抽出. / 小川原光一; 田邉康史; 倉爪亮; 長谷川勉.

In: 日本ロボット学会誌, Vol. 29, No. 1, 15.01.2011, p. 67-76.

Research output: Contribution to journalArticle

@article{bca6de26ee874475a90736cddd7dce3a,
title = "Partly Locality Sensitive Hashing を用いた時系列データからの高頻度パターン抽出",
abstract = "Frequent patterns in time series data are useful clues to learn previously unknown events in an unsupervised way. In this paper, we propose a method for detecting frequent patterns in long time series data efficiently. The major contribution of the paper is two-fold: (1) Partly Locality Sensitive Hashing (PLSH) is proposed to find frequent patterns efficiently and (2) the problem of finding consecutive time frames that have a large number of frequent patterns is formulated as a combinatorial optimization problem which is solved via Dynamic Programming (DP) in polynomial time O (N 1+1/α) thanks to PLSH where N is the total amount of data. The proposed method was evaluated by detecting frequent whole body motions in a video sequence as well as by detecting frequent everyday manipulation tasks in motion capture data.",
author = "光一 小川原 and 康史 田邉 and 亮 倉爪 and 勉 長谷川",
year = "2011",
month = "1",
day = "15",
doi = "10.7210/jrsj.29.67",
language = "Japanese",
volume = "29",
pages = "67--76",
journal = "日本ロボット学会誌",
issn = "0289-1824",
publisher = "一般社団法人 日本ロボット学会",
number = "1",

}

TY - JOUR

T1 - Partly Locality Sensitive Hashing を用いた時系列データからの高頻度パターン抽出

AU - 小川原, 光一

AU - 田邉, 康史

AU - 倉爪, 亮

AU - 長谷川, 勉

PY - 2011/1/15

Y1 - 2011/1/15

N2 - Frequent patterns in time series data are useful clues to learn previously unknown events in an unsupervised way. In this paper, we propose a method for detecting frequent patterns in long time series data efficiently. The major contribution of the paper is two-fold: (1) Partly Locality Sensitive Hashing (PLSH) is proposed to find frequent patterns efficiently and (2) the problem of finding consecutive time frames that have a large number of frequent patterns is formulated as a combinatorial optimization problem which is solved via Dynamic Programming (DP) in polynomial time O (N 1+1/α) thanks to PLSH where N is the total amount of data. The proposed method was evaluated by detecting frequent whole body motions in a video sequence as well as by detecting frequent everyday manipulation tasks in motion capture data.

AB - Frequent patterns in time series data are useful clues to learn previously unknown events in an unsupervised way. In this paper, we propose a method for detecting frequent patterns in long time series data efficiently. The major contribution of the paper is two-fold: (1) Partly Locality Sensitive Hashing (PLSH) is proposed to find frequent patterns efficiently and (2) the problem of finding consecutive time frames that have a large number of frequent patterns is formulated as a combinatorial optimization problem which is solved via Dynamic Programming (DP) in polynomial time O (N 1+1/α) thanks to PLSH where N is the total amount of data. The proposed method was evaluated by detecting frequent whole body motions in a video sequence as well as by detecting frequent everyday manipulation tasks in motion capture data.

U2 - 10.7210/jrsj.29.67

DO - 10.7210/jrsj.29.67

M3 - 記事

VL - 29

SP - 67

EP - 76

JO - 日本ロボット学会誌

JF - 日本ロボット学会誌

SN - 0289-1824

IS - 1

ER -