Mawadatul Maulidah
Universitas Bina Sarana Informatika, Jakarta

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Komparasi Evaluasi Kinerja Siswa Belajar dengan Mengggunakan Algoritma Machine Learning Elin Panca Saputra; Mawadatul Maulidah; Nadiyah Hidayati; Andi Saryoko
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 6, No 4 (2022): Oktober 2022
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v6i4.4786

Abstract

In our current study, we are doing a comparison of several algorithms that we have tested, namely in searching for the accuracy level of learning performance in students, the problem of this research is how to get the results of excellent generalization abilities so that a higher accuracy value is obtained. Our goal is to get the best-performing accuracy level results and then to identify features that can affect student learning performance. From the results of the algorithm that we have tested, four of them are Naïve Bayes, Support Vectore Machine, Neural Network and KNN contained in machine learning. The results of the four algorithms for the Naïve Bayes algorithm have an accuracy value of 96.30%, the Support Vectore Machine algorithm has an accuracy of 98.70%, and the Naural Network algorithm has an accuracy of 99.50% and the last one with the KNN algorithm produces an accuracy of 94.80%. it can be concluded that using the Neural Network algorithm is an algorithm with the best performance than using other algorithms in evaluating student learning performance, besides that the Neural Network can be used as an excellent alternative to be used as predictions, especially in the field of education.
Komparasi Machine Learning Berbasis Pso Untuk Prediksi Tingkat Keberhasilan Belajar Berbasis E-Learning Elin Panca Saputra; Siti Nurajizah; Mawadatul Maulidah; Nadiyah Hidayati; Taufik Rahman
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 10 No 2: April 2023
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.20231026469

Abstract

Perkembangan bidang teknologi memiliki aspek perkermbangan yang begitu cepat. penelitian kami memiliki tujuan untuk mentransmisikan sebuah pengetahuan tentang machine learning yang telah menjadi begitu popular digunakan hingga saat ini, pada penelitian ini bagaimana mendapatkan fitur seleksi atribut dan mendapatkan hasil prediksi dari pembelajaran pada Universitas atau lembaga Pendidikan yang menerapkan belajar dengan metode pembelajaran jarak jauh ataupun e-learning di era pandemic ini. Permasalahan pada penelitian ini yaitu jumlah atribut pada data dapat mengurangi akurasi, maka dari percobaan dengan beberapa algoritma pada machine learning kami mencoba menerapkan Particle Swarm Optimizatio(PSO) untuk meningkatkan akurasi yang lebih tinggi. Maka dari itu dapat disimpulkan penerapan menggunakan algoritma Naïve Bayes(NB) berbasis PSO mendapatkan hasil kenerja dengan bobot sebesar 94.40% dan angka AUC sebesar 94.50%, berikutnya Algoritma Support Vectore Machine(SVM) Berbasis PSO dengan hasil kinerja akurasi sebesar 88.20 dan nilai AUC seberar 91.10%, dan Artificial Neural Network(NN) berbasis Particle Swarm Optimizatio(PSO) menghasilkan skor hasil kinerja akurasi dengan bobot 99.20% dan nilai akurasi sebesar 98.50%, maka Artificial Neural Network(NN)  berbasis PSO memiliki keunggulan lebih besar dari pada algoritma naïve bayer berbasis PSO dan Support Vector Machine(SVM) dengan PSO. Sedangkan atribut yang mempunyai pengaruh menentukan dari algoritma tersebut pada tingkat akurasi adalah Practice Questions, Quizzes, Midterm exams, dan Final exams. terbukti dari penelitian-penelitian kami yang sebelumnya maka algoritma neural network berbasis PSO memang memiliki keunggulan yang begitu baik. Karena ANN merupakan metode yang memiliki perhitungan yang membangun beberapa unit pada saat pemrosesan berdasarkan koneksitas yang saling berhubungan, metode ANN dengan akurasi prediksi dapat menjadi sebuah alat yang efisien dan baik untuk penelitian estimasi dan klasifikasi dalam bidang pendidikan. Abstract The development of the field of technology has a very fast development aspect. our research has the aim of transmitting knowledge about machine learning which has become so popularly used until now, in this study how to get attribute selection features and get predictive results from learning at universities or educational institutions that apply learning by distance learning methods or e-learning. -Learning in this pandemic era. The problem in this study is that the number of attributes in the data can reduce accuracy, so from experiments with several yahoos on machine learning, we tried to apply Particle Swarm Optimizatio (PSO) to increase higher accuracy. Then the application key using the PSO-based Naïve Bayes (NB) algorithm can get performance results with a weight of 94.40% and an-AUC number of 94.50%, then the PSO-based Support Vectore Machine (SVM) Algorithm with a performance result of 88.20 and an AUC value of 91.10%, and Artificial Neural Network-(NN) based on Particle Swarm Optimizatio (PSO) produces an accuracy performance score with a weight of 99.20% and an accuracy value of 98.50%. Support Vector Machine (SVM) with PSO. While the attributes that have an influence to determine the algorithm on the level of accuracy are Practice Questions, Quizzes, Mid-Semester Exams, and Final Exams. it is evident from our previous studies that the PSO-based neural network algorithm does have a very good advantage. based on ANN is a method that has calculations that build several units of interconnected connectivity, the ANN method with predictive accuracy can be an efficient and good tool for forecasting and classification research in the field of education.