228,562 Hits in 6.1 sec

Algorithm Selection on a Meta Level [article]

Alexander Tornede, Lukas Gehring, Tanja Tornede, Marcel Wever, Eyke Hüllermeier
2021 arXiv   pre-print
Although most of these methods are based on machine learning, surprisingly little work has been done on meta learning, that is, on taking advantage of the complementarity of existing algorithm selection  ...  In this paper, we introduce the problem of meta algorithm selection, which essentially asks for the best way to combine a given set of algorithm selectors.  ...  Acknowledgements This work was partially supported by the German Research Foundation (DFG) within the Collaborative Research Center "On-The-Fly Computing" (SFB 901/3 project no. 160364472) and the German  ... 
arXiv:2107.09414v1 fatcat:4dygwremnndlxjot6l5re7s7ba

One-at-a-time: A Meta-Learning Recommender-System for Recommendation-Algorithm Selection on Micro Level [article]

Andrew Collins, Dominika Tkaczyk, Joeran Beel
2018 arXiv   pre-print
In this paper, we propose a meta-learning-based approach to recommendation, which aims to select the best algorithm for each user-item pair.  ...  Choosing a single algorithm based on overall evaluation results is not optimal.  ...  Micro-level meta-learners attempt to select the best algorithm for every instance in a dataset, or every single recommendation request on a given platform.  ... 
arXiv:1805.12118v3 fatcat:qyggmtvjebe7vnrjau5u7etntm

Towards Meta-Algorithm Selection [article]

Alexander Tornede, Marcel Wever, Eyke Hüllermeier
2020 arXiv   pre-print
We elaborate on consequences of applying AS on a meta-level and identify possible problems. Empirically, we show that meta-algorithm-selection can indeed prove beneficial in some cases.  ...  As an algorithm selector is again an algorithm solving a specific problem, the idea of algorithm selection could also be applied to AS algorithms, leading to a meta-AS approach: Given an instance, the  ...  Acknowledgments and Disclosure of Funding This work was partially supported by the German Research Foundation (DFG) within the Collaborative Research Center "On-The-Fly Computing" (SFB 901/3 project no  ... 
arXiv:2011.08784v1 fatcat:rj3m3nmqknhdbhfazvshcjocsa

u-cf2vec: Representation Learning for Personalized Algorithm Selection in Recommender Systems [article]

Tomas Sousa-Pereira, Tiago Cunha, Carlos Soares
2021 arXiv   pre-print
Meta Learning (MtL) has been used to choose the best algorithm for a given problem. Meta Learning is usually applied to select algorithms for a whole dataset.  ...  Adapting it to select the to the algorithm for a single user in a RS involves several challenges.  ...  We used a set of 5 meta models to try to solve the algorithm selection problem on the user level.  ... 
arXiv:2103.05673v1 fatcat:gmb24gv6ajfbrb3zykkbcsyqcy

A Novel Approach to Recommendation Algorithm Selection using Meta-Learning

Andrew Collins, Dominika Tkaczyk, Jöran Beel
2018 Irish Conference on Artificial Intelligence and Cognitive Science  
In this paper, we propose a meta-learning-based approach to recommendation, which aims to select the best algorithm for each user-item pair.  ...  Choosing a single algorithm based on overall evaluation results is not optimal.  ...  Ekstrand and Riedl [10] propose a mid-level meta-learner; they attempt to select the best algorithm for subsets of data in a dataset.  ... 
dblp:conf/aics/CollinsTB18 fatcat:wde7uhlh7rhghdfio7dg27uemq

Meta-Learning for Periodic Algorithm Selection in Time-Changing Data

Andre Luis Debiaso Rossi, Andre C.P.L.F. Carvalho, Carlos Soares
2012 2012 Brazilian Symposium on Neural Networks  
In this paper we present a meta-learning approach for periodic algorithm selection when data distribution may change over time.  ...  When users have to choose a learning algorithm to induce a model for a given dataset, a common practice is to select an algorithm whose bias suits the data distribution.  ...  META-LEARNING FOR PERIODIC ALGORITHMS SELECTION This paper presents a new approach for periodic algorithms selection in non-stationary environments based on meta-learning, named MetaStream.  ... 
doi:10.1109/sbrn.2012.50 dblp:conf/sbrn/RossiCS12 fatcat:mch43ipynvd75jadjuc32ehq2q

A Two-Leveled Web Service Path Re-PlanningTechnique

Shih Chien Chou
2012 International Journal on Web Service Computing  
This paper proposes a two-leveled path re-planning technique (TLPRP), which offers the following features: (a) TLPRP is composed of both meta and physical levels.  ...  (b) The physical level re-planning algorithm possesses the ability of web service path composition.  ...  With the meta level QoS criteria and the solutions identified by applying a system design process, the meta re-planning algorithm can select one optimal solution according to the meta level QoS criteria  ... 
doi:10.5121/ijwsc.2012.3402 fatcat:t36j55bffjggvj3zsd6vg2psuu


Ricardo Vilalta, Youssef Drissi
2012 Artificial Intelligence Review  
We find that, despite different views and research lines, a question remains constant: how can we exploit knowledge about learning (i.e. meta-knowledge) to improve the performance of learning algorithms  ...  The second part provides a survey of meta-learning as reported by the machine-learning literature.  ...  Note 1 A learning task is a 3-tuple, (F, m, ) , comprising a target concept F , a training-set size m, and a sample distribution from which the examples in the training set are drawn.  ... 
doi:10.1023/a:1019956318069 fatcat:jomateq3svejzhumcbxf6s6uwi

Improved Dataset Characterisation for Meta-learning [chapter]

Yonghong Peng, Peter A. Flach, Carlos Soares, Pavel Brazdil
2002 Lecture Notes in Computer Science  
This paper presents new measures, based on the induced decision tree, to characterise datasets for meta-learning in order to select appropriate learning algorithms.  ...  Totally 15 measures are proposed to describe the structure of a decision tree.  ...  The second is to select a group of learning algorithms including not only the best algorithm but also the algorithms that are not significantly worse than the best one.  ... 
doi:10.1007/3-540-36182-0_14 fatcat:sklygj6r45cezlw7ibjdjwdpqu

Contrasting meta-learning and hyper-heuristic research: the role of evolutionary algorithms

Gisele L. Pappa, Gabriela Ochoa, Matthew R. Hyde, Alex A. Freitas, John Woodward, Jerry Swan
2013 Genetic Programming and Evolvable Machines  
In this paper we first provide a historical perspective on automated algorithm design, and then we discuss similarities and differences between meta-learning in the field of supervised machine learning  ...  This discussion focuses on the dimensions of the problem space, the algorithm space and the performance measure, as well as clarifying important issues related to different levels of automation and generality  ...  Meta-evolutionary algorithms (such as meta-genetic algorithms and metagenetic programming) use evolutionary algorithms in a nested fashion, which occurs on two levels.  ... 
doi:10.1007/s10710-013-9186-9 fatcat:rxkljeappbea5m6bjmwahuqf24

Selecting Collaborative Filtering Algorithms Using Metalearning [chapter]

Tiago Cunha, Carlos Soares, André C. P. L. F. de Carvalho
2016 Lecture Notes in Computer Science  
The approach is tested on a set of Matrix Factorization algorithms and a collection of real-world Collaborative Filtering datasets.  ...  The algorithm selection problem is formulated as classification tasks, where the target attribute is the best Matrix Factorization algorithm, according to each metric.  ...  Ciência e a Tecnologia as part of project UID/EEA/50014/2013.  ... 
doi:10.1007/978-3-319-46227-1_25 fatcat:o44s3dsanfcotgmpffehv4rcea

Effect of Metalearning on Feature Selection Employment

Silvia Nunes das Dôres, Carlos Soares, Duncan D. A. Ruiz
2017 European Conference on Principles of Data Mining and Knowledge Discovery  
There are multiple methods for feature selection, with varying impact and computational cost. Therefore, choosing the right method for a given data set is important.  ...  This issue is relevant because a wrong decision may imply additional processing, when FS is unnecessarily applied, or in a loss of performance, when not used in a problem for which it is appropriate.  ...  Ciência e a Tecnologia (Portuguese Foundation for Science and Technology) as part of project UID/EEA/50014/2013.  ... 
dblp:conf/pkdd/DoresSR17 fatcat:g7q37gnrcbcj5pr2psdc5qf4uq

Metareasoning Structures, Problems, and Modes for Multiagent Systems: A Survey

Samuel T. Langlois, Oghenetekevwe Akoroda, Estefany Carrillo, Jeffrey W. Herrmann, Shapour Azarm, Huan Xu, Michael Otte
2020 IEEE Access  
metareasoning policy that the meta-level uses to select the task allocation algorithm that the object-level uses.)  ...  In the example mentioned earlier, the meta-level action is selecting the planning algorithm, the object level includes the planning algorithm, and the ground level includes actions such as moving through  ... 
doi:10.1109/access.2020.3028751 fatcat:tsbq3fwfwzcx5ed2e535unfava

A Distance-Weighted Class-Homogeneous Neighbourhood Ratio for Algorithm Selection

Haofei Chen, Ya Liu, Japnit Kaur Ahuja, Daren Ler
2020 Asian Conference on Machine Learning  
In this paper, we introduce a new form of meta-feature that is based on a distance-weighted class-homogeneous neighbourhood ratio to facilitate algorithm selection.  ...  Finally, in this paper, we provide a new perspective on landmarkers, such that a landmarker corresponds to a tuple (algorithm, metric), and propose the idea of a new family of meta-features.  ...  Each base-level algorithm a in the algorithm space A is trained over D to produce a base-level model, a(D) = m a .  ... 
dblp:conf/acml/ChenLAL20 fatcat:ad6qg5sluza5zip4ymqfhcigmm

A Review of Meta-level Learning in the Context of Multi-component, Multi-level Evolving Prediction Systems [article]

Abbas Raza Ali, Marcin Budka, Bogdan Gabrys
2020 arXiv   pre-print
It becomes a challenge in the presence of numerous configurations of learning algorithms on massive amounts of data.  ...  The techniques that are commonly used by experts are based on a trial and error approach evaluating and comparing a number of possible solutions against each other, using their prior experience on a specific  ...  The k-NN algorithm was used at the Meta-level to select the best candidate algorithm for a new dataset.  ... 
arXiv:2007.10818v1 fatcat:4jzeippeyjbxfdbza3h5222xey
« Previous Showing results 1 — 15 out of 228,562 results