The Experts below are selected from a list of 133080 Experts worldwide ranked by ideXlab platform
Ioannis Vlahavas - One of the best experts on this subject based on the ideXlab platform.
-
large scale online semantic indexing of biomedical articles via an Ensemble of multi label classification models
Journal of Biomedical Semantics, 2017Co-Authors: Yannis Papanikolaou, Grigorios Tsoumakas, Manos Laliotis, Nikos Markantonatos, Ioannis VlahavasAbstract:In this paper we present the approach that we employed to deal with large scale multi-label semantic indexing of biomedical papers. This work was mainly implemented within the context of the BioASQ challenge (2013–2017), a challenge concerned with biomedical semantic indexing and question answering. Our main contribution is a MUlti-Label Ensemble Method (MULE) that incorporates a McNemar statistical significance test in order to validate the combination of the constituent machine learning algorithms. Some secondary contributions include a study on the temporal aspects of the BioASQ corpus (observations apply also to the BioASQ’s super-set, the PubMed articles collection) and the proper parametrization of the algorithms used to deal with this challenging classification task. The Ensemble Method that we developed is compared to other approaches in experimental scenarios with subsets of the BioASQ corpus giving positive results. In our participation in the BioASQ challenge we obtained the first place in 2013 and the second place in the four following years, steadily outperforming MTI, the indexing system of the National Library of Medicine (NLM). The results of our experimental comparisons, suggest that employing a statistical significance test to validate the Ensemble Method’s choices, is the optimal approach for ensembling multi-label classifiers, especially in contexts with many rare labels.
-
large scale online semantic indexing of biomedical articles via an Ensemble of multi label classification models
arXiv: Machine Learning, 2017Co-Authors: Yannis Papanikolaou, Grigorios Tsoumakas, Manos Laliotis, Nikos Markantonatos, Ioannis VlahavasAbstract:Background: In this paper we present the approaches and Methods employed in order to deal with a large scale multi-label semantic indexing task of biomedical papers. This work was mainly implemented within the context of the BioASQ challenge of 2014. Methods: The main contribution of this work is a multi-label Ensemble Method that incorporates a McNemar statistical significance test in order to validate the combination of the constituent machine learning algorithms. Some secondary contributions include a study on the temporal aspects of the BioASQ corpus (observations apply also to the BioASQ's super-set, the PubMed articles collection) and the proper adaptation of the algorithms used to deal with this challenging classification task. Results: The Ensemble Method we developed is compared to other approaches in experimental scenarios with subsets of the BioASQ corpus giving positive results. During the BioASQ 2014 challenge we obtained the first place during the first batch and the third in the two following batches. Our success in the BioASQ challenge proved that a fully automated machine-learning approach, which does not implement any heuristics and rule-based approaches, can be highly competitive and outperform other approaches in similar challenging contexts.
-
random k labelsets an Ensemble Method for multilabel classification
European conference on Machine Learning, 2007Co-Authors: Grigorios Tsoumakas, Ioannis VlahavasAbstract:This paper proposes an Ensemble Method for multilabel classification. The RAndom k-labELsets (RAKEL) algorithm constructs each member of the Ensemble by considering a small random subset of labels and learning a single-label classifier for the prediction of each element in the powerset of this subset. In this way, the proposed algorithm aims to take into account label correlations using single-label classifiers that are applied on subtasks with manageable number of labels and adequate number of examples per label. Experimental results on common multilabel domains involving protein, document and scene classification show that better performance can be achieved compared to popular multilabel classification approaches.
Yihua Lin - One of the best experts on this subject based on the ideXlab platform.
-
multi step wind speed forecasting based on numerical simulations and an optimized stochastic Ensemble Method
Applied Energy, 2019Co-Authors: Jing Zhao, Jianzhou Wang, Zhenhai Guo, Yanling Guo, Wantao Lin, Yihua LinAbstract:Abstract At present, a single-valued deterministic simulation Method is preferred choice for numerical wind speed forecasts. However, it remains difficult to meet the actual needs of both wind farms and grid systems, mainly owing to unavoidable uncertainties. The development of skilled numerical forecasting Methods has become a critical issue and major challenge, and new capabilities and strategies for mitigating uncertainties in wind data derived from numerical models are highly sought after. On this topic, our study develops an improved Ensemble Method for day-ahead forecast of local wind speeds. The proposed Method constructs an optimized system based on Ensemble simulations of weather research and forecasting model, a Markov stochastic process, and an improved induced ordered weighted average approach that combines gray relationships with an evolutionary algorithm. The original contributions are concluded as: (i) using a Markov stochastic process, the observed information can be transferred to the Ensemble system, which contributes to the accuracy improvement; and (ii) the optimized induced ordered weighted average model, with a member selection process, is a new data-driven Ensemble Method for numerical wind speed forecasting. Simulation indicates that the proposed Method effectively reduces the uncertainties of numerical simulations, and performs better than other models. The simulation also shows that an Ensemble with fewer members may generate better results than using a combination of all single members. This study is of great significance for both theoretical research and real applications for numerical wind speed forecasts at local sites.
Yanfei Li - One of the best experts on this subject based on the ideXlab platform.
-
big multi step wind speed forecasting model based on secondary decomposition Ensemble Method and error correction algorithm
Energy Conversion and Management, 2018Co-Authors: Zhu Duan, Yanfei LiAbstract:Abstract Wind power is one of the most promising powers. Wind speed forecasting can eliminate the harmful effect caused by the intermittent and fluctuation of wind power, and big multi-step forecasting can provide more time for the power grid to be adjusted. To achieve the high-precision big multi-step forecasting, a novel hybrid model named as the WD-SampEn-VMD-MadaBoost-BFGS-WF is proposed in the study, which consisting of three main modeling steps including the secondary decomposition, the Ensemble Method and the error correction. The detail of the proposed model is given as follows: (a) wind speed series are decomposed by the WD (Wavelet Decomposition) to obtain wind speed subseries. The SampEn (Sample Entropy) algorithm is used to estimate the unpredictability of these wind speed subseries. The most unpredictable subseries will be decomposed secondarily by the VMD (Variational Mode Decomposition); (b) the subseries are proceeded by the MAdaBoost (Modified AdaBoost.RT) with the BFGS (Broyden–Fletcher–Goldfarb–Shanno Quasi-Newton Back Propagation) neuron network to obtain forecasting subseries; (c) all of the forecasting subseries will be combined with the original subseries to form the combined wind speed series, which will be further proceeded by the WF (Wavelet Filter) to obtain the corrected forecasting series from the point of the frequency domain; (d) the corrected forecasting series are reconstructed to get the final forecasting series. To validate the effectiveness of the proposed model, several forecasting cases are provided in the study. The result indicates that the proposed model has satisfactory forecasting performance in the big multi-step extremely strong simulating wind speed forecasting.
Jing Zhao - One of the best experts on this subject based on the ideXlab platform.
-
multi step wind speed forecasting based on numerical simulations and an optimized stochastic Ensemble Method
Applied Energy, 2019Co-Authors: Jing Zhao, Jianzhou Wang, Zhenhai Guo, Yanling Guo, Wantao Lin, Yihua LinAbstract:Abstract At present, a single-valued deterministic simulation Method is preferred choice for numerical wind speed forecasts. However, it remains difficult to meet the actual needs of both wind farms and grid systems, mainly owing to unavoidable uncertainties. The development of skilled numerical forecasting Methods has become a critical issue and major challenge, and new capabilities and strategies for mitigating uncertainties in wind data derived from numerical models are highly sought after. On this topic, our study develops an improved Ensemble Method for day-ahead forecast of local wind speeds. The proposed Method constructs an optimized system based on Ensemble simulations of weather research and forecasting model, a Markov stochastic process, and an improved induced ordered weighted average approach that combines gray relationships with an evolutionary algorithm. The original contributions are concluded as: (i) using a Markov stochastic process, the observed information can be transferred to the Ensemble system, which contributes to the accuracy improvement; and (ii) the optimized induced ordered weighted average model, with a member selection process, is a new data-driven Ensemble Method for numerical wind speed forecasting. Simulation indicates that the proposed Method effectively reduces the uncertainties of numerical simulations, and performs better than other models. The simulation also shows that an Ensemble with fewer members may generate better results than using a combination of all single members. This study is of great significance for both theoretical research and real applications for numerical wind speed forecasts at local sites.
Zhu Duan - One of the best experts on this subject based on the ideXlab platform.
-
big multi step wind speed forecasting model based on secondary decomposition Ensemble Method and error correction algorithm
Energy Conversion and Management, 2018Co-Authors: Zhu Duan, Yanfei LiAbstract:Abstract Wind power is one of the most promising powers. Wind speed forecasting can eliminate the harmful effect caused by the intermittent and fluctuation of wind power, and big multi-step forecasting can provide more time for the power grid to be adjusted. To achieve the high-precision big multi-step forecasting, a novel hybrid model named as the WD-SampEn-VMD-MadaBoost-BFGS-WF is proposed in the study, which consisting of three main modeling steps including the secondary decomposition, the Ensemble Method and the error correction. The detail of the proposed model is given as follows: (a) wind speed series are decomposed by the WD (Wavelet Decomposition) to obtain wind speed subseries. The SampEn (Sample Entropy) algorithm is used to estimate the unpredictability of these wind speed subseries. The most unpredictable subseries will be decomposed secondarily by the VMD (Variational Mode Decomposition); (b) the subseries are proceeded by the MAdaBoost (Modified AdaBoost.RT) with the BFGS (Broyden–Fletcher–Goldfarb–Shanno Quasi-Newton Back Propagation) neuron network to obtain forecasting subseries; (c) all of the forecasting subseries will be combined with the original subseries to form the combined wind speed series, which will be further proceeded by the WF (Wavelet Filter) to obtain the corrected forecasting series from the point of the frequency domain; (d) the corrected forecasting series are reconstructed to get the final forecasting series. To validate the effectiveness of the proposed model, several forecasting cases are provided in the study. The result indicates that the proposed model has satisfactory forecasting performance in the big multi-step extremely strong simulating wind speed forecasting.
