Open AccessPrediction Models with Machine Learning Against Student Success in Online Learning
Author(s) -
Yennimar,
Rohni Endetta Manihuruk,
Etis Landya Br Hotang
Publication year - 2021
Publication title -
sinkron
Language(s) - English
Resource type - Journals
eISSN - 2541-2019
pISSN - 2541-044X
DOI - 10.33395/sinkron.v6i1.11095
Subject(s) - artificial intelligence , computer science , artificial neural network , machine learning , academic year , online learning , covid-19 , mathematics education , test (biology) , psychology , multimedia , medicine , disease , pathology , infectious disease (medical specialty) , paleontology , biology
The learning system during the Covid-19 pandemic was carried out online, online learning had a negative impact and a positive impact. The impact given can affect the success of student learning. The success of learning is the main thing that must be achieved by students. From the success of learning, it can be seen that the online learning process is going well or not. To determine the success rate of online learning, testing is carried out by applying a neural network algorithm. Neural network algorithms are used because they can solve complex problems related to prediction. This research is expected to help lecturers or campus parties to create better online learning. In this study using student grade data for Academic Year 2018/2019 and Academic Year 2019/2020, data testing using Rapidminer software and operator cross validation. In testing the Academic Year 2018/2019 and Academic Year 2019/2020 value data using 700 training cycles, 0.4 momentum, 0.2 learning rate and hidden layer 2. The level of accuracy obtained in the 2018/2019 student grade data is 95, 55% and Academic Year 2019/2020 which is 93.17%. From the test results, it was found that the accuracy rate of Academic Year 2018/2019 is higher than Academic Year 2019/2020, so the success rate in Academic Year 2018/2019 before the pandemic is better than the success rate in Academic Year 2019/2020 after the pandemic.