Evaluating Feature Selection Algorithms

Molina, Luis Carlos; Belanche, Lluís; Nebot, Àngela

doi:10.1007/3-540-36079-4_19

Cited by 16 publications

(3 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To learn a performance‐influence model, we use a domain independent combination of multivariate regression and forward feature selection : we use multivariate regression to determine the coefficients of the different terms of a given model (eg, 3.01 for term π 1 ) and forward feature selection to select the terms that are iteratively added to the model . The general idea of the learning procedure is to start with a simple model and to iteratively expand it to a more complex but also more accurate model.…”

Section: Performance‐influence Modelsmentioning

confidence: 99%

Performance‐influence models of multigrid methods: A case study on triangular grids

Grebhahn

Rodrigo

Siegmund

et al. 2017

Concurrency and Computation

View full text Add to dashboard Cite

Summary Multigrid methods are among the most efficient algorithms for solving discretized partial differential equations. Typically, a multigrid system offers various configuration options to tune performance for different applications and hardware platforms. However, knowing the best performing configuration in advance is difficult, because measuring all multigrid system variants is costly. Instead of direct measurements, we use machine learning to predict the performance of the variants. Selecting a representative set of configurations for learning is nontrivial, although, but key to prediction accuracy. We investigate different sampling strategies to determine the tradeoff between accuracy and measurement effort. In a nutshell, we learn a performance‐influence model that captures the influences of configuration options and their interactions on the time to perform a multigrid iteration and relate this to existing domain knowledge. In an experiment on a multigrid system working on triangular grids, we found that combining pair‐wise sampling with the D‐Optimal experimental design for selecting a learning set yields the most accurate predictions. After measuring less than 1 % of all variants, we were able to predict the performance of all variants with an accuracy of 95.9 %. Furthermore, we were able to verify almost all knowledge on the performance behavior of multigrid methods provided by 2 experts.

show abstract

Section: Performance‐influence Modelsmentioning

confidence: 99%

Performance‐influence models of multigrid methods: A case study on triangular grids

Grebhahn

Rodrigo

Siegmund

et al. 2017

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…They efficiently traverse the space of subsets, by adding and deleting basis functions and use an evaluation measure that directs the search into areas of increased performance. Considering [7,8,9], in order to characterize a heuristic search problem one must define the following: 1) initial state of the search; 2) available state-transition operators; 3) search strategy; 4) evaluation measure; 5) termination condition. In the subset selection approach for polynomial regression, typically the initial states are models that correspond to the empty subset, the subset with only the intercept term in it, full subset of all the defined basis functions, or a randomly chosen subset; the typical state-transition operators are addition and deletion of a basis function; the typical search strategy is the hill climbing [2,3,7] which in combination with the empty subset initial state and the addition operator becomes SFS but in combination with the full subset initial state and the deletion operator becomes SBS; the classical evaluation measures are the statistical significance tests [1], however, currently two other strategies predominate: employment of complexity penalization criteria (e.g.…”

Section: Subset Selectionmentioning

confidence: 99%

A comparison of subset selection and adaptive basis function construction for polynomial regression model building

Jēkabsons¹,

Lavendels²

2009

Scientific Journal of Riga Technical University. Computer Sciences

View full text Add to dashboard Cite

A comparison of subset selection and adaptive basis function construction for polynomial regression model buildingThe approach of subset selection in polynomial regression model building assumes that the chosen fixed full set of predefined basis functions contains a subset that is sufficient to describe the target relation sufficiently well. However, in most cases the necessary set of basis functions is not known and needs to be guessed - a potentially non-trivial (and long) trial and error process. In our previous research we considered an approach for polynomial regression model building which is different from the subset selection - letting the regression model building method itself construct the basis functions necessary for creating a model of arbitrary complexity without restricting oneself to the basis functions of a predefined full model. The approach is titled Adaptive Basis Function Construction (ABFC). In the present paper we compare the two approaches for polynomial regression model building - subset selection and ABFC - both theoretically and empirically in terms of their underlying principles, computational complexity, and predictive performance. Additionally in empirical evaluations the ABFC is compared also to two other well-known regression modelling methods - Locally Weighted Polynomials and Multivariate Adaptive Regression Splines.

show abstract

“…The sequential floating forward selection ( SFFS ) developed by Pudil et al (1994) was chosen as the second method, because it needs only one parameter to be specified and has been shown in several evaluation studies to be as good as or better than other selection techniques (Reunanen 2003; Molina et al 2002).…”

mentioning

confidence: 99%

Ipra—enhancing the Sensing Abilities of Ambient Intelligence

Schultheis

2007

Computational Intelligence

View full text Add to dashboard Cite

The ability of ambient intelligence to serve a person could be considerably enhanced, if intelligent devices would take into account the person's current cognitive and affective state. One primary source for state information are physiological measures. However, obtaining state information requires concertedly employing a multitude of different techniques. To facilitate the use of physiological signals as sources of state information and thereby allow improving ambient intelligence, we developed IPRA, an integrated pattern recognition approach which (a) employs all techniques necessary to appropriately analyze physiological data, (b) is broadly applicable, and (c) is easy to use even for nonexperts. This approach together with results of its first application are presented in this contribution.

show abstract

Evaluating Feature Selection Algorithms

Cited by 16 publications

References 5 publications

Performance‐influence models of multigrid methods: A case study on triangular grids

Performance‐influence models of multigrid methods: A case study on triangular grids

A comparison of subset selection and adaptive basis function construction for polynomial regression model building

Ipra—enhancing the Sensing Abilities of Ambient Intelligence

Contact Info

Product

Resources

About