PENGGUNAAN DEEP LEARNING DALAM ANALISIS POLA KESULITAN BELAJAR MATEMATIKA DI MAS AL HIKMAH BANDAR LAMPUNG

Penelitian Kualitatif Deskriptif

Abstract

This research is motivated by the low mathematics achievement of Indonesian students, as reflected in the results of the 2022 PISA (Programme for International Student Assessment) and pre-research findings at MAS Al Hikmah Bandar Lampung, where 10th-grade students experienced difficulties in understanding basic concepts, carrying out problem-solving procedures, and connecting concepts. The purpose of this study was to analyze patterns of students' mathematics learning difficulties and identify the potential application of deep learning technology to support the learning process. This study used a descriptive qualitative method with 20 10th-grade students and mathematics teachers as supporting informants. Data were collected through questionnaires, interviews, and documentation, then analyzed using the Miles & Huberman interactive model. The results showed that students still faced conceptual and procedural difficulties, while teachers assessed that emotional factors were not dominant compared to weak mastery of basic concepts. The questionnaire results revealed that understanding of deep learning was still low (mean 2.26), barriers to implementation were high (mean 1.68), but optimism regarding its benefits was quite high (mean 3.10), especially in analyzing student learning error patterns. These findings emphasize the importance of utilizing deep learning as an innovative way to analyze learning difficulties, supported by facilities, teacher training, and gradual implementation to make mathematics learning more effective and on-target.