Abstract:In image classification, feature combination is often used to combine the merits of multiple complementary features and improve the classification accuracy compared with one single feature. Existing feature combination algorithms, e.g., multiple kernel learning, usually determine the weights of features based on the optimization with respect to some classifier-dependent objective function. These algorithms are often computationally expensive, and in some cases are found to perform no better than simple baselines. In this paper, we solve the feature combination problem from a totally different perspective. Our algorithm is based on the simple idea of combining only base kernels suitable to be combined. Since the very aim of feature combination is to obtain the highest possible classification accuracy, we measure the combination suitableness of two base kernels by the maximum possible cross-validation accuracy of their combined kernel. By regarding the pairwise suitableness as the kernel adjacency, we obtain a weighted graph of all base kernels and find that the base kernels suitable to be combined correspond to a cluster in the graph. We then use the dominant sets algorithm to find the cluster and determine the weights of base kernels automatically. In this way, we transform the kernel combination problem into a clustering one. Our algorithm can be implemented in parallel easily and the running time can be adjusted based on available memory to a large extent. In experiments on several data sets, our algorithm generates comparable classification accuracy with the state of the art.

Multiple Classifiers Combination by Clustering and Selection.

Multiple Classifiers Combination Based on Specialists' FIelds

A New Multiple Classifiers Combination Algorithm

Combination of Multiple Classifier Using Feature Space Partitioning

Decision by Maximum of Posterior Probability Average with Weights: a Method of Multiple Classifiers Combination

Research and application of universal combination rule of multiple classifiers system

Multiple Decision Template Classifier Combination

Dynamic Selection and Circulating Combination for Multiple Classifier Systems

Combination of Multiple Nearest Neighbor Classifiers Based on Feature Subset Clustering Method

Multiple clustering and selecting algorithms with combining strategy for selective clustering ensemble

Multiple Classifier Combination for Hyperspectral Remote Sensing Image Classification

Combination of Heterogeneous Multiple Classifiers Based on Evidence Theory

K-means Selective Cluster Ensembles based on Multiple Feature Subsets

Support Vector Machine for Multiple Feature Classifcation

Applying Multiple Classifier Combination by Clustering and Selection to SoftMan's Perception System

Decision combination in multiple classifier systems

Multiple Classifier Fusion Based on Testing Sample Pairs

Feature Combination Via Clustering

Learning the Classifier Combination for Image Classification

Evaluating Classifier Combination in Object Classification

Isomerous Multiple Classifier Ensemble Method With Svm And Kmp