LAUSR

The main objective of this study is to predict student academic resilience based on academic emotions in studying numeration and science under online learning. Many researchers have analyzed student academic resilience and online learning. Unfortunately, only a few similar research topics focus on numeration and science. 191 students at a university in Central Java Province have been randomly selected as research samples. Academic resilience is classified into three groups: low, medium, and high. The academic emotions were measured using three indicators: class-related emotions, learning-related emotions, and test emotions. This study uses an artificial neural network (ANN) to obtain predictive values. The results indicate that the level of academic resilience and academic emotion in numeration and science under online learning is in the medium category. The results also show that the relative error provides a fairly small percentage, namely 19.7% at the training stage and 25% at the testing stage. This refers to the prediction results having a good level of accuracy. Predictive estimation results also indicate that class-related emotions are predicted to be the aspect that has the most crucial impact on students’ academic resilience, in which the normalized importance value is 100.0%. It is followed by the aspect of learning-related emotions (65.0%) and test emotions (24.3%). The implication is that the aspect of class-related emotions should get better attention from lecturers and students so that students can increase their chances of getting a good level of academic resilience in numeration and science.