KLASIFIKASI JENIS BURUNG BERDASARKAN SUARA MENGGUNAKAN ALGORITME SUPPORT VECTOR MACHINE

Abstract

Birds are one of the animals that need to be protected because of some extinct species. To solve this problem, conservation are needed because not all bird habitats can be reached by humans. Bird sound was one of the unique markers to indicate the existence of bird life. With technology, information, and communication, human could identify bird species by its sound. The classification process based sound is one of the problems that can be solved by Machine Learning. The noise reduction process is needed to remove noise that is not needed like wind and rain. Noise Reduction using Wavelet Transform result is sound signal more cleaner with PSNR value 4,2890. The classification of bird species is based on sound using one of the machine learning algorithms, Support Vector Machine (SVM) with Heavy Tailed RBF kernel with ratio 7:3. The accuracy produced from this study was 77,00% obtained from the use of parameters namely c = 10.000 and epsilon = 1.00E-09.