scholarly journals Comparative Analysis of Facial Affect Detection Algorithms

2020 ◽  
Author(s):  
Ashin Marin Thomas ◽  
Myounghoon Jeon
Keyword(s):  
Comparative Analysis ◽  
Facial Affect ◽  
Detection Algorithms ◽  
Affect Detection

scholarly journals СРАВНИТЕЛЬНЫЙ АНАЛИЗ АЛГОРИТМОВ ОБНАРУЖЕНИЯ НЕИЗВЕСТНЫХ СИГНАЛОВ С УЧЁТОМ СОВОКУПНОСТИ ПОКАЗАТЕЛЕЙ КАЧЕСТВА

2018 ◽  
pp. 67-74
Author(s):  
Валерий Михайлович Безрук ◽  
Станислав Андреевич Иваненко
Keyword(s):  
Comparative Analysis ◽  
Signal Detection ◽  
Quality Indicators ◽  
Performance Indicators ◽  
A Priori ◽  
Pareto Optimal ◽  
Detection Algorithms ◽  
A Priori Uncertainty ◽  
Unknown Signal

The subject of this article is the problem of detecting unknown signals in conditions of high a priori uncertainty, which can occur during the determination of unoccupied frequency channels in cognitive networks. It should be noted that various sources of radio emissions work on the air, which in turn complicates the solution of the problem of detection, since it is impossible to say what kind of signal will be received. Most existing algorithms require information about the signals for their operation. In practice, it is not possible to have such data on all sources of radio emission due to their diversity. The goal of the article is to study non-classical signal detection algorithms in conditions of high a priori uncertainty, when there is information only about noise, and signals are unknown.  The task: to conduct a comparative analysis of unknown signal detection algorithms based on a set of quality indicators and to determine the set of Pareto-optimal detection algorithms, as well as to identify the best algorithm for a set of quality indicators.  The method of statistical modeling of detection algorithms on samples of real signals and noise is performed. As a result, we obtained estimates of speed of work and quality of signal detection for a number of different variants of unknown signal detection algorithms. Possible variants of implementation of the detectors were summarized in the table. These variants were formed taking into account the dimension of the DPF sample and the number of implementations on which the decision is made. A comparative analysis of different types of detection algorithms is carried out taking into account the set of performance indicators and the quality of signal detection. It should be noted that the values of quality indicators of detection of unknown signals and performance indicators of the algorithms are related and contradictory. Conclusions. A multicriteria selection of a subset of Pareto-optimal variants is performed. Using the conditional preference criterion, the only preferred variant of the algorithm for detecting unknown signals is selected from the Pareto subset. The results of the research can be used in automated radio monitoring in cognitive radio networks

Comparative Analysis of Drift Detection Based Adaptive Ensemble Model with Different Drift Detection Techniques

2021 ◽  
Vol 23 (06) ◽  
pp. 49-55
Author(s):  
Sanjeev Kumar ◽  
◽  
Ravendra Singh ◽  
Keyword(s):  
Data Mining ◽  
Comparative Analysis ◽  
False Positive ◽  
Opinion Mining ◽  
Concept Drift ◽  
Ensemble Classifier ◽  
Stream Data ◽  
Detection Techniques ◽  
Stream Data Mining ◽  
Detection Algorithms

Stream data mining is a popular research area these days. The concept drift detection and drift handling are the biggest challenges of stream data mining. Several drift detection algorithms have been developed which can accurately detect various drifts but have the problem of false-positive drift detection. The false-positive drift detection leads to the performance degradation of the classifier because of unnecessary training in between analyses. Classifier ensemble has shown its efficiency for drift detection, drift handling, and classification. But the ensemble classifiers could not detect the exact position of drift occurrence, so it has to update itself at some fixed interval, which leads to an unnecessary computational burden on the system. Combining the drift detection algorithm with an ensemble classifier can improve the performance and also solve the problems of false-positive drift detection and unnecessary updating of the ensemble classifier. In this paper, a model is proposed that creates a weighted adaptive ensemble classifier by updating it only when a drift detection signal is given by the used drift detection method. The proposed model is evaluated on text-based stream data for sentiment analysis and opinion mining with multiple drift detection algorithms and with multiple classification algorithms as base classifiers for the ensemble. A comparative analysis has been done, and the results have shown the efficiency of the proposed models.

Sign in / Sign up

Export Citation Format

Share Document