Random pairwise shapelets forest: an effective classifier for time series
被引:8
|
作者:
Yuan, Jidong
论文数: 0引用数: 0
h-index: 0
机构:
Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R ChinaBeijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China
Yuan, Jidong
[1
]
Shi, Mohan
论文数: 0引用数: 0
h-index: 0
机构:
Beijing Jingdong 360 Degree E Commerce Co Ltd, Beijing, Peoples R ChinaBeijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China
Shi, Mohan
[2
]
Wang, Zhihai
论文数: 0引用数: 0
h-index: 0
机构:
Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R ChinaBeijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China
Wang, Zhihai
[1
]
Liu, Haiyang
论文数: 0引用数: 0
h-index: 0
机构:
Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R ChinaBeijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China
Liu, Haiyang
[1
]
Li, Jinyang
论文数: 0引用数: 0
h-index: 0
机构:
Univ Hong Kong, Dept Comp Sci, Pok Fu Lam, Hong Kong, Peoples R ChinaBeijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China
Li, Jinyang
[3
]
机构:
[1] Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China
[2] Beijing Jingdong 360 Degree E Commerce Co Ltd, Beijing, Peoples R China
[3] Univ Hong Kong, Dept Comp Sci, Pok Fu Lam, Hong Kong, Peoples R China
Time series classification;
Pairwise shapelets;
Random forest;
Decomposed mean decrease impurity;
REPRESENTATION;
SIMILARITY;
FEATURES;
D O I:
10.1007/s10115-021-01630-z
中图分类号:
TP18 [人工智能理论];
学科分类号:
081104 ;
0812 ;
0835 ;
1405 ;
摘要:
Shapelet is a discriminative subsequence of time series. An advanced shapelet-based method is to embed shapelet into the accurate and fast random forest. However, there are several limitations. First, random shapelet forest requires a large training cost for split threshold searching. Second, a single shapelet provides limited information for only one branch of the decision tree, resulting in insufficient accuracy. Third, the randomized ensemble decreases comprehensibility. For that, this paper presents Random Pairwise Shapelets Forest (RPSF). RPSF combines a pair of shapelets from different classes to construct random forest. It omits threshold searching to be more efficient, includes more information about each node of the forest to be more effective. Moreover, a discriminability measure, Decomposed Mean Decrease Impurity, is proposed to identify the influential region for each class. Extensive experiments show that RPSF is competitive compared with other methods, while it improves the training speed of shapelet-based forest.