PEMODELAN SISTEM REKOMENDASI CERDAS MENGGUNAKAN HYBRID DEEP LEARNING

Dewi, Kadek Cahya and Ciptayani, Putu Indah (2022) PEMODELAN SISTEM REKOMENDASI CERDAS MENGGUNAKAN HYBRID DEEP LEARNING. Jurnal Sistem Informasi dan Sains Teknologi, 4 (2). pp. 1-7. ISSN 2684-8260

Text PLT-Ac5-Pemodelan Sistem Reomendasi Cerdas.pdf Download (527kB)

Abstract

The current trend of technological development is intelligent systems. However, there is no research that combines the two deep learning methods on recommendation algorithms. It is important to model intelligent recommendation systems using hybrid deep learning This study aimed to obtain an optimal hybrid deep learning model on an intelligent recommendation system. This study used an experimental research approach. Data collection techniques included observation, online search and record datasets. The research stages consisted of: (a) literature review, (b) observation and online search, (c) modeling (d) prototyping and (e) testing. The model was a hybrid deep learning model consisting of two layers, namely the Self Organizing Map (SOM) layer and the Recurrent Neural Network (RNN) layer. This research used the Python programming language in the prototyping stage. Some libraries modules in python namely numpy, pandas, tensorflow, hard, torch, sklearn were used. Program tested with dataset from kaggle.com. The test results managed to improve performance by increasing the accuracy to 100%. It can be concluded that the SOM-RNN model can improve the performance of the intelligent recommendation system

Actions (login required)

View Item