An Heuristic Feature Selection Algorithm to Evaluate Academic Performance of Students

Ajibade, Samuel-Soma M.; Ahmad, Nor Ezzaty; Shamsuddin, Siti Mariyam

doi:10.1109/icsgrc.2019.8837067

Cited by 29 publications

(14 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…"The efficient decisions, anomaly, and non-repetition characteristics of the logistic map have the potential to increase wide range of population, and also enhance the functioning of converging at global optimum, avoiding untimely convergence" [8]. It is shown in equation (4).…”

Section: Inertia Weightmentioning

confidence: 99%

“…IOP Publishing doi:10.1088/1742-6596/2250/1/012016 2 "The total space of exploring is made up of all possible subsets of capabilities, indicating that the degree of the request space is 2n, where n is the number of actual capabilities" [4]. "Metaheuristicsbased techniques have recently gotten a lot of attention because of their great result in tackling the feature selection challenges and also because of their future global search ability" [2].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Utilizing Logistic Map to Enhance the Population Diversity of PSO

Ajibade

Chweya

Ogunbolu

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Particle swarm optimization (PSO) is a high-quality, nature-inspired global optimization algorithm that can be applied to a variety of real-world optimization problems. PSO, on the other hand, has some flaws, such as slow convergence, premature convergence, and the ability to become stuck at local optimum solutions. This research aims to address the issue of population diversity in the PSO search process, which leads to premature convergence. As a result, in this study, a method is introduced to PSO in order to avoid early stagnation, which leads to premature convergence. A chaotic dynamic weight particle swarm optimization (CTPSOA) is proposed, in which a chaotic logistic map is delivered to increase the population range within the PSO search technique by editing the inertia weight of PSO to avoid premature convergence. This study also looks into the overall performance and viability of the proposed CTPSOA as a set of function selection rules for solving optimization issues. There are eight traditional benchmark functions that are used to assess the overall results and obtain the accuracy of the proposed (CTPSOA) algorithms when compared to a few other meta-heuristics optimization rules. The test results reveal that the CTPSOA outperforms other meta-heuristics algorithms in solving optimization problems by 85% and has established itself as a reliable and superior metaheuristics algorithm for feature selection.

show abstract

Section: Inertia Weightmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Utilizing Logistic Map to Enhance the Population Diversity of PSO

Ajibade

Chweya

Ogunbolu

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…This paper using six feature selection algorithms have been tested before, there are Cfs subset eval [15], Chi squared attribute eval [16], filtered attribute eval [17], gain ratio attribute eval [18], principal components [19], and relief attribute eval [20]. This paper also uses 15 different classification algorithms that have been tested through educational datasets, specifically Bayes net, Naïve Bayes, Naive Bayes updateable, multilayer perceptron, simple logistic, SMO, decision tree, JRip, OneR, PART, decision stump, J48, random forest, random tree, and REP tree [21]- [23].…”

Section: Feature Selection Algorithm and Classificationmentioning

confidence: 99%

Comparison of the feature selection algorithm in educational data mining

Triayudi

Fitri

2021

TELKOMNIKA

View full text Add to dashboard Cite

Student academic accomplishment is the foremost focus of every educational institution. In developing student achievement in educational institutions, the researchers finally created a new research area, namely educational data mining (EDM). How the feature selection (FS) algorithm works is by removing unrelated data from educational datasets; therefore, this algorithm can improve the classification performance managed in EDM techniques. This research presents an analysis of the performance of the FS algorithm from the student dataset. The results received from other FS algorithms and classifiers will help other researchers to gain some best combination regarding FS algorithms and the classification. Selecting features that are relevant for student forecast models is a sensitive problem to stakeholders in education because they must make decisions based on the results of the prediction models. For the future, our paper seeks to play a decisive part while developing quality concerning education, as well as guiding different researchers in conducting educational interventions.

show abstract

“…Feature Selection (FS) algorithm thereby enhances the classifier's performance used by EDM technology and removes inappropriate data from educational databases. This convolution based classification model shows very high classification performance, and the C5.0 classification is known for its precision and accuracy [18]. Besides, the importance of variables for each taxonomic model is analyzedto predict based on Support Vector machine [19].Moreover deep learning features are intent to improve the performance accuracy, success as an instructor is based on students 'perceptions, mainly on students' course of interest study [20].…”

Section: Related Workmentioning

confidence: 99%

Psychotic Motivation for Improving Student Performance Based On Pattern Learner Features Using Deep Neural Classifier for Bipolar Disorder Students

Peerbasha¹,

Surputheen²

2021

cibg

View full text Add to dashboard Cite

Bipolar disorder is a depressive fact that makes manic illness pressures in young ages due to the non-intensive nature of brain functions, energy levels, mood-outs, and health disorders. These abnormalities may affect student performance under the learning strategies of students. Improvement of bipolar disorder affected student performance needs more data analysis forums that lead to high dimensional nature of features. The problem is that nonrelation feature analysis depends on the nature of student fitness that creates low prediction during classifications for students' motivation. To resolve this problem, a Psychotic motivation is proposed for improving student performance based on Pattern Learner Features (PLF) using Intra Segment Recurrent Deep Neural Network (ISRDNN) for bipolar disorder students. The proposed system makes student academic data's with physical fitness data collection progressive approach to predict important features to classify the result. Bipolar Disorder Influence Rate (BDIR) is usedto spill the progressive student defectives and the learning capabilities for classification result. With Intra Segment Activation Function (ISAF), the recurrent neural network is optimized to classify the result. This classifier improved the student's academic performance based on psychological motivation recommendations. Results prove that the accuracy of the proposed system produces high results compared to the previous system.

show abstract

An Heuristic Feature Selection Algorithm to Evaluate Academic Performance of Students

Cited by 29 publications

References 9 publications

Utilizing Logistic Map to Enhance the Population Diversity of PSO

Utilizing Logistic Map to Enhance the Population Diversity of PSO

Comparison of the feature selection algorithm in educational data mining

Psychotic Motivation for Improving Student Performance Based On Pattern Learner Features Using Deep Neural Classifier for Bipolar Disorder Students

Contact Info

Product

Resources

About