A Non-Parametric Method to Determine Basic Probability Assignment Based on Kernel Density Estimation

Qin, Bowen; Xiao, Fuyuan

doi:10.1109/access.2018.2883513

Cited by 33 publications

(15 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, a new non-parametric method based on KDE has been proposed to determine BPA [20,21]. Inspire by the idea, a score-mapping method for forensic is presented in this paper, in which KDE and evidence theory are used to determine the confidence of whether two face image pairs to compare are from the same source or from the difference source.…”

Section: Proposed Methodsmentioning

confidence: 99%

Forensic face recognition based on KDE and evidence theory

Wen

2021

MATEC Web Conf.

View full text Add to dashboard Cite

Forensic face recognition (FFR) has been studied in recent years in forensic science. Given an automatic face recognition system, output scores of the system are used to describe the similarity of face image pairs, but not suitable for forensics. In this study, a score-mapping model based on kernel density estimation (KDE) and evidence theory is proposed. First, KDE was used to generate probability density function (PDF) for each dimensional feature vector of face image pairs. Then, the PDFs could be utilized to determine separately the basic probability assignment (BPA) of supporting the prosecution hypothesis and the defence hypothesis. Finally, the BPAs of each feature were combined by Dempster’s rule to get the final BPA, which reflects the strength of evidence support. The experimental results demonstrate that compared with the classic KDE-based likelihood ratio method, the proposed method has a better performance in terms of accuracy, sensitivity and specificity.

show abstract

Section: Proposed Methodsmentioning

confidence: 99%

Forensic face recognition based on KDE and evidence theory

Wen

2021

MATEC Web Conf.

View full text Add to dashboard Cite

show abstract

“…The aforementioned statistical tolerance method is applicable to the processing of Gaussian vibration data but cannot be used for the induction of non-Gaussian data. To solve the induction problem of measured non-Gaussian vibration data, the Johnson or Bootstrap method is used [11], [12], [21]. Before Step 2 above, it is necessary to first convert non-Gaussian data into data that approximately obey a normal distribution by using the Johnson or Bootstrap method.…”

Section: ) Estimation Of the Upper Tolerance Limit Of The Psdmentioning

confidence: 99%

Kernel Density Estimation Based Gaussian and Non-Gaussian Random Vibration Data Induction for High-Speed Train Equipment

Wang

Deng

Wang³

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Because general statistics tolerance is not applicable to the induction of non-Gaussian vibration data and the methods for converting non-Gaussian data into Gaussian data are not always effective and can increase the estimation error, a novel kernel density estimation method in which induction is carried out on power spectral density data for the measured vibration of high-speed trains is proposed in this paper. First, data belonging to the same population of power spectral density are merged into the same feature sample. Then, the probability density function of all power spectral density values at the first frequency point is calculated through the kernel density estimation method, and the upper-limit estimate of all power spectral density values under the set quantile is obtained. This process is repeated, and the upper limit values of the power spectral density values at all frequency points can be obtained to convert the measured acceleration data to the acceleration power spectral density spectrum of the vibration test. Engineering examples are used to verify the proposed method. For the same Gaussian power spectral density data, the relative error between the root mean squares of the power spectral density spectrum obtained from induction by the kernel density estimation method and the statistics tolerance is 0.155% ∼ 1.55%; for the non-Gaussian power spectral density data, the acceleration power spectral density spectrum of the non-Gaussian vibration can be obtained with the induction by the kernel density estimation method. The proposed kernel density estimation method satisfies the induction requirements for the measured Gaussian and non-Gaussian vibration data of high-speed trains with two different distributions, and its induction results have very good universality and estimation accuracy.

show abstract

“…First, a unified methodology for assigning basic probability (BP) to each proposition has not been fully developed. Many researchers have proposed different methods for assigning BP as follows: non-parametric [24], kernel distribution [25], normal distribution [26], k-mean clustering [27], and fuzzy theory [10], [28]. In the past few years, some studies have used a generalized fuzzy number (GFN) [29], a trapezoidal fuzzy number [30], and a triangular fuzzy number (TFN) [17], [19], [22], [31] methods to improve classification performance.…”

Section: Introductionmentioning

confidence: 99%

Evidence-Theoretic Reentry Target Classification Using Radar: A Fuzzy Logic Approach

et al. 2021

View full text Add to dashboard Cite

This study focuses on the reentry target classification and fuses target features based on the generalized evidence theory. The features are extensively investigated, and the ballistic factor and length of the high-resolution range profile are selected. The evidence theory is advantageous for solving feature fusion, representing uncertainty, and is widely used in defense applications. However, determining the generalized basic probability assignment (GBPA) and dealing with uncertainty is a matter that requires further improvement. In this paper, we propose a new method to determine GBPA using uncertainty with time-series radar data. First, the samples of each known class are encoded as a generalized fuzzy number (GFN), and the power set comprising the frame of discernment (FOD) is calculated from the GFN and each intersection area. Subsequently, the test samples with uncertainty are encoded as triangular fuzzy numbers, reflecting the mean and standard deviation of a Kalman filter. Finally, the firing strength between the model and the input is calculated as the degree of support for the class hypothesis, which is used to determine the GBPA. The proposed algorithm is compared with the existing methods and exhibits high classification accuracy and a short classification time without leakage. In experiments with various input uncertainties, the results demonstrate that our method can effectively reflect the input uncertainty and determine the GBPA.

show abstract

A Non-Parametric Method to Determine Basic Probability Assignment Based on Kernel Density Estimation

Cited by 33 publications

References 47 publications

Forensic face recognition based on KDE and evidence theory

Forensic face recognition based on KDE and evidence theory

Kernel Density Estimation Based Gaussian and Non-Gaussian Random Vibration Data Induction for High-Speed Train Equipment

Evidence-Theoretic Reentry Target Classification Using Radar: A Fuzzy Logic Approach

Contact Info

Product

Resources

About