Combining Pattern Recognition Modalities at the Sensor Level Via Kernel Fusion

Multiple Classifier Systems

Sulimova

Windridge

et al. 2009

Self Cite

Abstract. We consider the problem of multi-modal pattern recognition under the assumption that a kernel-based approach is applicable within each particular modality. The Cartesian product of the linear spaces into which the respective kernels embed the output scales of single sensors is employed as an appropriate joint scale corresponding to the idea of combining modalities at the sensor level. This contrasts with the commonly adopted method of combining classifiers inferred from each specific modality. However, a significant risk in combining linear spaces is that of overfitting. To address this, we set out a stochastic method for encompassing modal-selectivity that is intrinsic to (that is to say, theoretically contiguous with) the selected kernel-based pattern-recognition approach. The principle of kernel selectivity supervision is then applied to the problem of signature verification by fusing several on-line and off-line kernels into a complete training and verification technique.

Section: The Continuous Training Technique With Supervised Selectivitymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Supervised Selective Combining Pattern Recognition Modalities and Its Application to Signature Verification by Fusing On-Line and Off-Line Kernels

Multiple Classifier Systems

Sulimova

Windridge

et al. 2009

Self Cite

“…This fusion may be performed at the early or late stage. In the former case of early fusion [1,2], the growing overall dimensionality of the object representation with increasing the number of modalities can be reduced by incorporating some form of modalityselection within the final classification procedure [3,4], thereby eliminating the danger of over-fitting. Such modality-selectivity is correlated with the generalization performance of the training process, so that, if performed ideally, the recognition system user is free to include object-representation modalities without constraint.…”

Section: Introductionmentioning

confidence: 99%

“…One such algorithm is given in [2] under the assumption that a kernel-based methodology is utilized to obtain a recognition rule for each particular modality, for which a discriminant hyperplane is specified in the linear space associated with each modality. In this case, the kernel trick [11,12] transforms the problem of combining diverse modalities with missing data into that of appropriately treating blanks in the modality-specific kernel matrices when fusing them into a unified matrix.…”

Section: Introductionmentioning

confidence: 99%

A Modified Neutral Point Method for Kernel-Based Fusion of Pattern-Recognition Modalities with Incomplete Data Sets

Panov

Multiple Classifier Systems

Mottl

et al. 2011

Self Cite

Abstract. It is commonly the case in multi-modal pattern recognition that certain modality-specific object features are missing in the training set. We address here the missing data problem for kernel-based Support Vector Machines, in which each modality is represented by the respective kernel matrix over the set of training objects, such that the omission of a modality for some object manifests itself as a blank in the modalityspecific kernel matrix at the relevant position. We propose to fill the blank positions in the collection of training kernel matrices via a variant of the Neutral Point Substitution (NPS) method, where the term "neutral point" stands for the locus of points defined by the "neutral hyperplane" in the hypothetical linear space produced by the respective kernel. The current method crucially differs from the previously developed neutral point approach in that it is capable of treating missing data in the training set on the same basis as missing data in the test set. It is therefore of potentially much wider applicability. We evaluate the method on the Biosecure DS2 data set.

Supervised Selective Kernel Fusion for Membrane Protein Prediction

Pattern Recognition in Bioinformatics

Sulimova

Torshin

et al. 2014

Self Cite

Abstract. Membrane protein prediction is a significant classification problem, requiring the integration of data derived from different sources such as protein sequences, gene expression, protein interactions etc. A generalized probabilistic approach for combining different data sources via supervised selective kernel fusion was proposed in our previous papers. It includes, as particular cases, SVM, Lasso SVM, Elastic Net SVM and others. In this paper we apply a further instantiation of this approach, the Supervised Selective Support Kernel SVM and demonstrate that the proposed approach achieves the top-rank position among the selective kernel fusion variants on benchmark data for membrane protein prediction. The method differs from the previous approaches in that it naturally derives a subset of "support kernels" (analogous to support objects within SVMs), thereby allowing the memory-efficient exclusion of significant numbers of irrelevant kernel matrixes from a decision rule in a manner particularly suited to membrane protein prediction.