# Kernelized Supervised Dictionary Learning

Source: [Kernelized Supervised Dictionary Learning](<https://www.sota2.com/research/paper/68fa612c1e209131ce9f2f41>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T19:53:34.667Z

Authors: Mehrdad J. Gangeh, Ali Ghodsi, Mohamed S. Kamel

Published: 2012-07-10T20:52:46.000Z

## Abstract

In this paper, we propose supervised dictionary learning (SDL) by incorporating information on class labels into the learning of the dictionary. To this end, we propose to learn the dictionary in a space where the dependency between the signals and their corresponding labels is maximized. To maximize this dependency, the recently introduced Hilbert Schmidt independence criterion (HSIC) is used. One of the main advantages of this novel approach for SDL is that it can be easily kernelized by incorporating a kernel, particularly a data-derived kernel such as normalized compression distance, into the formulation. The learned dictionary is compact and the proposed approach is fast. We show that it outperforms other unsupervised and supervised dictionary learning approaches in the literature, using real-world data.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Image Classification on Extended YaleB 30% corrupted](<https://www.sota2.com/research/sota/image-classification-on-extended-yaleb-30-percent-corrupted.md>): Image Classification; Extended YaleB 30% corrupted; Accuracy: 78.9.

- [Face Recognition on Extended YaleB](<https://www.sota2.com/research/sota/face-recognition-on-extended-yaleb.md>): Face Recognition; Extended YaleB; Accuracy: 95.65.

- [Handwritten Digit Recognition on MNIST](<https://www.sota2.com/research/sota/handwritten-digit-recognition-on-mnist.md>): Handwritten Digit Recognition; MNIST; Accuracy: 94.43.

- [Digit Classification on USPS](<https://www.sota2.com/research/sota/digit-classification-on-usps.md>): Digit Classification; USPS; Accuracy: 96.81.

- [Face Recognition on Cropped YaleB](<https://www.sota2.com/research/sota/face-recognition-on-cropped-yaleb.md>): Face Recognition; Cropped YaleB; Accuracy: 87.

- [Handwritten Digit Recognition on ARDIS](<https://www.sota2.com/research/sota/handwritten-digit-recognition-on-ardis.md>): Handwritten Digit Recognition; ARDIS; Accuracy: 93.24.

- [Telugu OCR on UHTelPCC](<https://www.sota2.com/research/sota/telugu-ocr-on-uhtelpcc.md>): Telugu OCR; UHTelPCC; Accuracy: 90.43.

- [Telugu OCR on Banti](<https://www.sota2.com/research/sota/telugu-ocr-on-banti.md>): Telugu OCR; Banti; Accuracy: 71.9.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1207.2488>)

[PDF](<https://arxiv.org/pdf/1207.2488.pdf>)