Segment and Combine Approach for Non-parametric Time-Series Classification

Geurts, Pierre; Wehenkel, Louis

doi:10.1007/11564126_48

Cited by 17 publications

(6 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both methods have already proven useful in a number of applications. In particular, problems of very high dimensionality, like image classification problems (Marée et al, 2004), mass-spectrometry datasets (Geurts et al, 2005b), or time-series classification problems (Geurts and Wehenkel, 2005), make the Extra-Trees a first choice method due to its attractive computational performances. Also, the fact that totally randomized trees have a tree structure independent of the output variable has been exploited in the context of reinforcement learning where it ensures the convergence of the reinforcement learning algorithm and leads to a very efficient implementation (Ernst et al, 2005).…”

Section: Resultsmentioning

confidence: 99%

17 Extremely Randomized Trees

2021

Data Science for Supply Chain Forecasting

View full text Add to dashboard Cite

This paper proposes a new tree-based ensemble method for supervised classification and regression problems. It essentially consists of randomizing strongly both attribute and cut-point choice while splitting a tree node. In the extreme case, it builds totally randomized trees whose structures are independent of the output values of the learning sample. The strength of the randomization can be tuned to problem specifics by the appropriate choice of a parameter. We evaluate the robustness of the default choice of this parameter, and we also provide insight on how to adjust it in particular situations. Besides accuracy, the main strength of the resulting algorithm is computational efficiency. A bias/variance analysis of the Extra-Trees algorithm is also provided as well as a geometrical and a kernel characterization of the models induced.

show abstract

Section: Resultsmentioning

confidence: 99%

17 Extremely Randomized Trees

2021

Data Science for Supply Chain Forecasting

View full text Add to dashboard Cite

show abstract

“…The Extra‐Trees algorithm, a nonparametric tree‐based regression method based on an ensemble of decision trees, has already been applied in many fields such as environmental modeling [ Jung et al ., ] and water reservoir operation [ Castelletti et al ., ]. Because of its efficient computational performance in problems with very high dimensionality [ Geurts et al ., ; Geurts and Wehenkel , ], we selected the Extra‐Trees algorithm to determine the relationships between variables in cascade reservoir operation.…”

Section: Methodsmentioning

confidence: 99%

Multiobjective reservoir operating rules based on cascade reservoir input variable selection method

Yang

Guo

Liu

et al. 2017

Water Resources Research

View full text Add to dashboard Cite

The input variable selection in multiobjective cascade reservoir operation is an important and difficult task. To address this problem, this study proposes the cascade reservoir input variable selection (CIS) method that searches for the most valuable input variables for decision making in multiple‐objectivity cascade reservoir operations. From a case study of Hanjiang cascade reservoirs in China, we derive reservoir operating rules based on the combination of CIS and Gaussian radial basis functions (RBFs) methods and optimize the rules through Pareto‐archived dynamically dimensioned search (PA‐DDS) with two objectives: to maximize both power generation and water supply. We select the most effective input variables and evaluate their impacts on cascade reservoir operations. From the simulated trajectories of reservoir water level, power generation, and water supply, we analyze the multiobjective operating rules with several input variables. The results demonstrate that the CIS method performs well in the selection of input variables for the cascade reservoir operation, and the RBFs method can fully express the nonlinear operating rules for cascade reservoirs. We conclude that the CIS method is an effective and stable approach to identifying the most valuable information from a large number of candidate input variables. While the reservoir storage state is the most valuable information for the Hanjiang cascade reservoir multiobjective operation, the reservoir inflow is the most effective input variable for the single‐objective operation of Danjiangkou.

show abstract

“…The methods are effective to analyse time series segments with limited size. Additionally, segment time series can be seen as series representation to summary time series features in [19][20][21]. However, they just can be used to analyse single time series.…”

Section: Related Workmentioning

confidence: 99%