Penerapan Teknologi Deep Learning dalam Media Pembelajaran Interaktif di Madrasah Ibtidaiyah: (Studi Kasus MI Krandegan 1 Gandusari Trenggalek)

Madrasah Ibtidaiyah (MI) has a strategic role in shaping students' literacy, numeracy, and concept understanding skills from an early age. However, most MI still applies conventional learning methods that have minimal use of technology. Along with the advancement of Artificial Intelligence (AI), especially deep learning, there are opportunities to present more interactive and adaptive learning. This article discusses the application of interactive learning media based on Convolutional Neural Network (CNN) at MI Krandegan 1 Gandusari, Trenggalek. These media are designed to recognize images and sounds as part of thematic learning that support students' visual and auditory understanding. The research was conducted for four months involving 45 students and 8 teachers. The methods used include observation, system development, teacher training, and learning implementation. The results showed that CNN-based interactive media was able to increase students' concept understanding by 31%, from 59% to 90%. The teacher stated that this media makes it easier to deliver material and increases students' enthusiasm for learning. In addition, student involvement in the learning process increases significantly, accompanied by active support from parents. These findings indicate that deep learning technology has great potential in improving the quality of learning in MI. The application of CNN-based media can be an innovative solution to answer the challenges of basic education in the digital era, while encouraging the transformation of more effective and fun learning methods.