A new classifier based on information theoretic learning with unlabeled data

Jeong, Kyu-Hwa; Xu, Jing; Erdoğmuş, Deni̇z; Prı́ncipe, José C.

doi:10.1016/j.neunet.2005.06.018

Cited by 23 publications

(12 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As one of the criteria utilizing distributions, the ED between the distribution of transmitted symbol f D (d) and the equalizer output distribution f Y (y) is defined as (3) [3,6].…”

Section: Ed Criterion and Entropymentioning

confidence: 99%

“…For training of adaptive systems for medical diagnosis, the ED criterion has been successfully applied to distinguish biomedical datasets [6]. For finite impulse response (FIR) adaptive filter structures in impulsive noise environments, ED between the output distribution and a set of Dirac delta functions has been used as an efficient performance criterion taking advantage of the outlier-cutting effect of Gaussian kernel for output pairs and symbol-output pairs [7].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Function Analysis of the Euclidean Distance between Probability Distributions

Kim

2018

Entropy

View full text Add to dashboard Cite

Minimization of the Euclidean distance between output distribution and Dirac delta functions as a performance criterion is known to match the distribution of system output with delta functions. In the analysis of the algorithm developed based on that criterion and recursive gradient estimation, it is revealed in this paper that the minimization process of the cost function has two gradients with different functions; one that forces spreading of output samples and the other one that compels output samples to move close to symbol points. For investigation the two functions, each gradient is controlled separately through individual normalization of each gradient with their related input. From the analysis and experimental results, it is verified that one gradient is associated with the role of accelerating initial convergence speed by spreading output samples and the other gradient is related with lowering the minimum mean squared error (MSE) by pulling error samples close together.

show abstract

“…As one of the criteria utilizing distributions, the ED between the distribution of transmitted symbol f D (d) and the equalizer output distribution f Y (y) is defined as (3) [3,6].…”

Section: Ed Criterion and Entropymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Function Analysis of the Euclidean Distance between Probability Distributions

Kim

2018

Entropy

View full text Add to dashboard Cite

show abstract

“…Based on minimization of ED, the authors in [18] proposed using the ED criterion to train the adaptive system in order to match two PDFs from different output patterns, and successively applied it to classification problems with a real biomedical data set [13]. As a similar approach, [14] applied an ED minimization method to compensate for sensor drift in which the RBFN weights are readjusted even during the test phase, but only with the use of output data samples that were obtained during the training phase.…”

Section: Test Phase Readjustment For Sensor Drift Compensationmentioning

confidence: 99%

Signal Processing Techniques Based on Adaptive Radial Basis Function Networks for Chemical Sensor Arrays

Byun¹

2016

Journal of Sensor Science and Technology

View full text Add to dashboard Cite

The use of a chemical sensor array can help discriminate between chemicals when comparing one sample with another. The ability to classify pattern characteristics from relatively small pieces of information has led to growing interest in methods of sensor recognition. A variety of pattern recognition algorithms, including the adaptive radial basis function network (RBFN), may be applicable to gas and/ or odor classification. In this paper, we provide a broad review of approaches for various types of gas and/or odor identification techniques based on RBFN and drift compensation techniques caused by sensor poisoning and aging.

show abstract

“…As a robust ITL-type algorithm, the Euclidean distance minimization between PDFs has been introduced by Jeong et al and applied successfully to the classification problem with a real biomedical data set [10]. The researchers in [10] proposed to reuse the previously acquired training-phase output samples in the test phase so that the test-phase output PDF follows the training-phase output PDF. In the research [11] for blind equalization, the Euclidean distance minimization method is applied using signal power for blind equalization.…”

Section: Introductionmentioning

confidence: 99%

Communication Equalizer Algorithms with Decision Feedback based on Error Probability

Kim

Hwang²

2011

Journal of the Korea Academia-Industrial cooperation Society

View full text Add to dashboard Cite

For intersymbol interference (ISI) compensation from communication channels with multi-path fading and impulsive noise, a decision feedback equalizer algorithm that minimizes Euclidean distance of error probability is proposed. The Euclidean distance of error probability is defined as the quadratic distance between the probability error signal and Dirac-delta function. By minimizing the distance with respect to equalizer weight based on decision feedback structures, the proposed decision feedback algorithm has shown to have significant effect of residual ISI cancellation on severe multipath channels as well as robustness against impulsive noise.

show abstract

A new classifier based on information theoretic learning with unlabeled data

Cited by 23 publications

References 6 publications

Function Analysis of the Euclidean Distance between Probability Distributions

Function Analysis of the Euclidean Distance between Probability Distributions

Signal Processing Techniques Based on Adaptive Radial Basis Function Networks for Chemical Sensor Arrays

Communication Equalizer Algorithms with Decision Feedback based on Error Probability

Contact Info

Product

Resources

About