Corpus ID: 28322951

Filters, Wrappers and a Boosting-Based Hybrid for Feature Selection

@inproceedings{Das2001FiltersWA,
  title={Filters, Wrappers and a Boosting-Based Hybrid for Feature Selection},
  author={Sanmay Das},
  booktitle={ICML},
  year={2001}
}
Apparatus for holding a camera and flash unit so they are separated from one another, including a frame having an upper portion for holding the flash unit and a lower portion for holding the camera, and having a pair of laterally spaced handles near the lower portion. The handles are oriented so they can be held comfortably at slightly below eye level with the upper arms extending down and braced against the body and the forearms extending upwardly. One of the handles forms a palm pad lying… Expand

Figures, Tables, and Topics from this paper

Conditional feature sensitivity: a unifying view on active recognition and feature selection
TLDR
This paper proposes a unified perspective through conditional feature sensitivity analysis, taking into account both current context and feature interactions, and presents three treatment models and exploits their joint power for dealing with complex feature interactions. Expand
Ordinal Feature Selection for Iris and Palmprint Recognition
TLDR
The proposed LP formulation for ordinal feature selection is advantageous over existing feature selection methods, such as mRMR, ReliefF, Boosting, and Lasso for biometric recognition, reporting state-of-the-art accuracy on CASIA and PolyU databases. Expand
Ordinal Feature Selection for IRIS and Palm print Recognition
Ordinal measures have been demonstrated as an effective feature representation model for iris and palmprint recognition. However, ordinal measures are a genera l concept of image analysis andExpand
A Multi-Gestures Recognition System Based on Less sEMG Sensors
TLDR
This work attempted to recognize six commonly used handgestures with two-channel sensors, and the classification performance and calculation time of different algorithms are compared, and achieved the recognition accuracy of 91.93%. Expand
Adaptive visual sampling
TLDR
This thesis proposes an adaptive spatial sampling strategy framework to maintain statistical object models for real-time robust tracking under changing lighting conditions and presents three frameworks that build on previous methods to provide a more flexible and effective approach. Expand
Feature selection for low bit rate mobile augmented reality applications
TLDR
Novel relevance-based feature selection for low bit rate MAR applications, based on the entropy of the image content, entropy of extracted features and the Discrete Cosine Transformation (DCT) coefficients are proposed, which achieve superior retrieval performance under lowbit rate. Expand
Improving Performance and Applying Cascades in Visual Classification Master Thesis
Computer-based object classification has improved consistently over the last decade, but the performance of current computational schemes is still significantly lower than human classificationExpand
Automatic Bifurcation Detection in Coronary IVUS Sequences
TLDR
A fully automatic method which identifies every bifurcation in an intravascular ultrasound (IVUS) sequence, the corresponding frames, the angular orientation with respect to the IVUS acquisition, and the extension is presented. Expand
Facial Action Unit Recognition Using Filtered Local Binary Pattern Features with Bootstrapped and Weighted ECOC Classifiers
TLDR
Within the context face expression classification using the facial action coding system (FACS), the problem of detecting facial action units (AUs) is addressed and it is shown that both these sources of error can be reduced by enhancing ECOC through the application of bootstrapping and class-separability weighting. Expand
Feature Selection Methods for Boosted Crosspectral Face Recognition
TLDR
Three novel feature selection methods are proposed: Genuine segment score thresholding, d′-based thresholding and two Adaboost inspired methods to prune irrelevant information in encoded data and to improve performance of the Boosted LGPI technique. Expand
...
1
2
3
4
5
...

References

SHOWING 1-9 OF 9 REFERENCES
Toward Optimal Feature Selection
TLDR
An efficient algorithm for feature selection which computes an approximation to the optimal feature selection criterion is given, showing that the algorithm effectively handles datasets with a very large number of features. Expand
Wrappers for Performance Enhancements and Oblivious Decision Graphs.
TLDR
This doctoral dissertation concludes that repeated runs of five-fold cross-validation give a good tradeoff between bias and variance for the problem of model selection used in later chapters. Expand
Correlation-based Feature Selection for Machine Learning
TLDR
This thesis addresses the problem of feature selection for machine learning through a correlation based approach with CFS (Correlation based Feature Selection), an algorithm that couples this evaluation formula with an appropriate correlation measure and a heuristic search strategy. Expand
A Short Introduction to Boosting
TLDR
This short overview paper introduces the boosting algorithm AdaBoost, and explains the underlying theory of boosting, including an explanation of why boosting often does not suffer from overfitting as well as boosting’s relationship to support-vector machines. Expand
Irrelevant Features and the Subset Selection Problem
TLDR
A method for feature subset selection using cross-validation that is applicable to any induction algorithm is described, and experiments conducted with ID3 and C4.5 on artificial and real datasets are discussed. Expand
Induction of Selective Bayesian Classifiers
TLDR
This paper embeds the naive Bayesian induction scheme within an algorithm that carries out a greedy search through the space of features, hypothesize that this approach will improve asymptotic accuracy in domains that involve correlated features without reducing the rate of learning in ones that do not. Expand
Greedy Attribute Selection
TLDR
Experiments suggest hillclimbing in attribute space can yield substantial improvements in generalization performance, and a caching scheme is presented that makes attribute hillClimbing more practical computationally. Expand
Learning to remove Internet advertisements
TLDR
The experiments demonstrate that the inductive learning approach to browsing assistant is practical: the off-line training phase takes less than six minutes; on-line classification takes about 70 msec; and classification accuracy exceeds 97% given a modest set of training data. Expand
Fast Committee Machines for Regression and Classification
TLDR
This work shows how both classification and prediction error can be reduced by using boosting techniques to implement committee machines, using either classification trees or regression trees. Expand