Comparison on Classification of the Holy Quran Verses using MFCC and RQA
Annisa Khodista Syaka, Noor Akhmad Setiawan, Oyas Wahyunggoro · 2021
Classification of the Holy Quran verses identifies the verses of the Holy Quran based on several features in the corresponding acoustic wave. In the study, the acoustic waves use the Tarteel dataset, which has many readers' voices. For comparison purposes, two feature extraction methods are implemented: Mel Frequency Cepstrum Coefficients (MFCC) and Recurrence Quantification Analysis (RQA). Extraction results and Artificial Neural Network (ANN) classifier are used to create general and each chapter model. Then the two-way prediction is performed in order to evaluate the results. According to the experiment, data extraction using the MFCC method obtains better classification results than the RQA. The prediction value with MFCC reaches 77% of the true value, while RQA obtains 65%.