Filters








331 Hits in 4.2 sec

Transfer Learning, Soft Distance-Based Bias, and the Hierarchical BOA [article]

Martin Pelikan, Mark W. Hauschild, Pier Luca Lanzi
2012 arXiv   pre-print
An automated technique has recently been proposed to transfer learning in the hierarchical Bayesian optimization algorithm (hBOA) based on distance-based statistics.  ...  The technique enables practitioners to improve hBOA efficiency by collecting statistics from probabilistic models obtained in previous hBOA runs and using the obtained statistics to bias future hBOA runs  ...  The use of bias based on the results of other learning tasks is also commonplace in machine learning where it is referred to as inductive transfer or transfer learning [22, 23] .  ... 
arXiv:1203.5443v2 fatcat:j6x4vugp5zcfnf4p35cnyd3k7q

Transfer Learning, Soft Distance-Based Bias, and the Hierarchical BOA [chapter]

Martin Pelikan, Mark W. Hauschild, Pier Luca Lanzi
2012 Lecture Notes in Computer Science  
An automated technique has recently been proposed to transfer learning in the hierarchical Bayesian optimization algorithm (hBOA) based on distance-based statistics.  ...  Abstract An automated technique has recently been proposed to transfer learning in the hierarchical Bayesian optimization algorithm (hBOA) based on distance-based statistics.  ...  Acknowledgments This project was sponsored by the National Science Foundation under grants ECS-0547013 and IIS-1115352, and by the Univ. of Missouri-St.  ... 
doi:10.1007/978-3-642-32937-1_18 fatcat:lyn34n636jd5pnlzpto4tdqfkm

Distance-based bias in model-directed optimization of additively decomposable problems

Martin Pelikan, Mark W. Hauschild
2012 Proceedings of the fourteenth international conference on Genetic and evolutionary computation conference - GECCO '12  
Compared to other techniques for learning from experience put forward in the past, the proposed technique is both more practical and more broadly applicable.  ...  While the focus of the paper is on additively decomposable problems and the hierarchical Bayesian optimization algorithm, it should be straightforward to generalize the approach to other modeldirected  ...  Acknowledgments This project was sponsored by the National Science Foundation under grants ECS-0547013 and IIS-1115352, and by the Univ. of Missouri in St.  ... 
doi:10.1145/2330163.2330203 dblp:conf/gecco/PelikanH12 fatcat:4uwnf7355jdyvmuuhwdaizghry

Distance-Based Bias in Model-Directed Optimization of Additively Decomposable Problems [article]

Martin Pelikan, Mark W. Hauschild
2012 arXiv   pre-print
Compared to other techniques for learning from experience put forward in the past, the proposed technique is both more practical and more broadly applicable.  ...  While the focus of the paper is on additively decomposable problems and the hierarchical Bayesian optimization algorithm, it should be straightforward to generalize the approach to other model-directed  ...  Acknowledgments This project was sponsored by the National Science Foundation under grants ECS-0547013 and IIS-1115352, and by the University of Missouri in St.  ... 
arXiv:1201.2241v1 fatcat:hh2lhwt5tnh7revkh2zflwq5ci

Transforming Evolutionary Search into Higher-Level Evolutionary Search by Capturing Problem Structure

Rob Mills, Thomas Jansen, Richard A. Watson
2014 IEEE Transactions on Evolutionary Computation  
Recently, a small number of works make use of such ideas by learning problem structure and using this information in a particular manner: these works use the results of a simple search process in primitive  ...  We discuss strengths and limitations of the multi-scale search approach and point out how it can be developed further.  ...  Such search heuristics are called unbiased [56] and they are more reasonable than biased search heuristics as long as a specific search bias is not justified by problem-specific knowledge.  ... 
doi:10.1109/tevc.2014.2347702 fatcat:rnyejjamlvgovpkhj2axvqyvli

Computational Methods to Interpret and Integrate Metabolomic Data [chapter]

Feng Li, Jiangxin Wang, Lei Nie, Weiwen Zhang
2012 Metabolomics  
As a popular unsupervised learning method, Hierarchical cluster analysis (HCA) clusters the data to form a tree diagram or dendrogram which shows the relationships between samples (Ebbels, 2007) .  ...  Besides counting matches, other measures also consider their actual mass and intensity, such as the Euclidean distance, the probability-based matching (PBM), the normalized dot product (NDP), and a modified  ...  This book will provide the reader with summaries of the state-of-the-art of technologies and methodologies, especially in the data analysis and interpretation approaches, as well as give insights into  ... 
doi:10.5772/32517 fatcat:ety5vgnvfnaw5evchxhgbab3wy

Context Based Predictive Information

Yuval Shalev, Irad Ben-Gal
2019 Entropy  
We propose a new algorithm called the context-based predictive information (CBPI) for estimating the predictive information (PI) between time series, by utilizing a lossy compression algorithm.  ...  The advantage of this approach over existing methods resides in the case of sparse predictive information (SPI) conditions, where the ratio between the number of informative sequences to uninformative  ...  Acknowledgments: We would like to thank Hadar Levi Aharoni for her insightful comments regarding the information bottleneck.  ... 
doi:10.3390/e21070645 pmid:33267359 pmcid:PMC7515138 fatcat:iafgwsciqbdwtdapxa7mno6tba

Complex Activity Recognition Via Attribute Dynamics

Wei-Xin Li, Nuno Vasconcelos
2016 International Journal of Computer Vision  
The binary dynamic system (BDS) model is proposed to jointly learn the distribution and dynamics of activities in this space.  ...  A BDS learning algorithm, inspired by the popular dynamic texture, and a dissimilarity measure between BDSs, which generalizes the Binet-Cauchy kernel, are introduced.  ...  Learning a WAD Vocabulary Given the BC-KLD distance between BDSs, it is possible to learn a WAD dictionary from a BoAS P = {⇧ (i) } N i=1 , by applying Algorithm 2 as follows.  ... 
doi:10.1007/s11263-016-0918-1 fatcat:yceautxluja5tij2jg2r5dpa2e

A Survey on the Explainability of Supervised Machine Learning [article]

Nadia Burkart, Marco F. Huber
2020 arXiv   pre-print
This survey paper provides essential definitions, an overview of the different principles and methodologies of explainable Supervised Machine Learning (SML).  ...  We conduct a state-of-the-art survey that reviews past and recent explainable SML approaches and classifies them according to the introduced definitions.  ...  This way, they can decide what to transfer and when to transfer in order to create an optimized transfer learning setting.  ... 
arXiv:2011.07876v1 fatcat:ccquewit2jam3livk77l5ojnqq

Forecasting stock price using integrated artificial neural network and metaheuristic algorithms compared to time series models

Milad Shahvaroughi Farahani, Seyed Hossein Razavi Hajiagha
2021 Soft Computing - A Fusion of Foundations, Methodologies and Applications  
Then, we used genetic algorithms (GA) as a heuristic algorithm for feature selection and choosing the best and most related indicators.  ...  The statistical population of research have five most important and international indices which were S&P500, DAX, FTSE100, Nasdaq and DJI.  ...  Spiders use these vibrations to understand the distance and transfer it from member i to member j.  ... 
doi:10.1007/s00500-021-05775-5 pmid:33935586 pmcid:PMC8070984 fatcat:6qhxmla3pnfith65hvir6ntlwy

A Survey on the Explainability of Supervised Machine Learning

Nadia Burkart, Marco F. Huber
2021 The Journal of Artificial Intelligence Research  
This survey paper provides essential definitions, an overview of the different principles and methodologies of explainable Supervised Machine Learning (SML).  ...  We conduct a state-of-the-art survey that reviews past and recent explainable SML approaches and classifies them according to the introduced definitions.  ...  Acknowledgments This work is partially supported by the Ministry of Economic Affairs of the state Baden-Württemberg within the KI-Fortschrittszentrum "Lernende Systeme", Grant No. 036-170017.  ... 
doi:10.1613/jair.1.12228 fatcat:nd3hfatjknhexb5eabklk657ey

Comprehensive Taxonomies of Nature- and Bio-inspired Optimization: Inspiration versus Algorithmic Behavior, Critical Analysis and Recommendations [article]

Daniel Molina and Javier Poyatos and Javier Del Ser and Salvador García and Amir Hussain and Francisco Herrera
2020 arXiv   pre-print
This paper addresses this problem by proposing two comprehensive, principle-based taxonomies that allow researchers to organize existing and future algorithmic developments into well-defined categories  ...  , considering two different criteria: the source of inspiration and the behavior of each algorithm.  ...  Leaders and followers - A new metaheuristic to avoid the bias of accumulated information.  ... 
arXiv:2002.08136v2 fatcat:paeupscdt5hgzjtqzdtpcuucfq

Unmasking Clever Hans predictors and assessing what machines really learn

Sebastian Lapuschkin, Stephan Wäldchen, Alexander Binder, Grégoire Montavon, Wojciech Samek, Klaus-Robert Müller
2019 Nature Communications  
Here we apply recent techniques for explaining decisions of state-of-the-art learning machines and analyze various tasks from computer vision and arcade games.  ...  Furthermore, we propose our semi-automated Spectral Relevance Analysis that provides a practically effective way of characterizing and validating the behavior of nonlinear learning machines.  ...  hierarchical priors which simplify learning new concepts from previous experience.  ... 
doi:10.1038/s41467-019-08987-4 pmid:30858366 pmcid:PMC6411769 fatcat:2r376sf72revhhwqiagj7j25mm

Is this what the debate on rules was about?

Ulrike Hahn
2005 Behavioral and Brain Sciences  
In the present article, I assume that the rules versus similarity distinction can be understood in the same way in learning, reasoning, categorization, and language, and that a unified model for rules  ...  It is argued that this viewpoint allows adequate coverage of theory and empirical findings in learning, reasoning, categorization, and language, and also a reassessment of the objectives in research on  ...  Other heuristics and biases are ambiguous with respect to a Rule/Similarity classification, but not for any profound reason; for example, the belief bias could be Similarity or Rules, depending on how  ... 
doi:10.1017/s0140525x05340015 fatcat:dqfkizkq2baf5cojlprx2de2ae

The rules versus similarity distinction

Emmanuel M. Pothos
2005 Behavioral and Brain Sciences  
In the present article, I assume that the rules versus similarity distinction can be understood in the same way in learning, reasoning, categorization, and language, and that a unified model for rules  ...  It is argued that this viewpoint allows adequate coverage of theory and empirical findings in learning, reasoning, categorization, and language, and also a reassessment of the objectives in research on  ...  ACKNOWLEDGMENTS We thank Jay McClelland, David Plaut, and members of the Carnegie Mellon PDP group for useful discussion related to this commentary.  ... 
doi:10.1017/s0140525x05000014 fatcat:hbc6e5p25vbunaikg5h3kuau2e
« Previous Showing results 1 — 15 out of 331 results