KLASIFIKASI EMOSI MUSIK BERDASARKAN FITUR AUDIO DENGAN MENGGUNAKAN METODE FUZZY K NEAREST NEIGHBOR
| Main Author: | KURNIAWAN, REZA TRIADI |
|---|---|
| Format: | Thesis NonPeerReviewed Book |
| Bahasa: | eng |
| Terbitan: |
, 2019
|
| Subjects: | |
| Online Access: |
http://eprints.umm.ac.id/51572/1/PENDAHULUAN.pdf http://eprints.umm.ac.id/51572/42/BAB%20I.pdf http://eprints.umm.ac.id/51572/43/BAB%20II.pdf http://eprints.umm.ac.id/51572/44/BAB%20III.pdf http://eprints.umm.ac.id/51572/45/BAB%20IV.pdf http://eprints.umm.ac.id/51572/46/BAB%20V.pdf http://eprints.umm.ac.id/51572/47/POSTER.pdf http://eprints.umm.ac.id/51572/ |
Daftar Isi:
- Music is full of unique attributes, especially concerning auditory perceptions. Amplitude elements, frequency and duration cannot be discerned yet as music to human ears. Before those three elements processed by human brains interpretation so that they became timbre pitch dynamics and tempo. In the process of music interpretation in the fields of musicology, there is also the term “emotion” that’s often used. Every music listener have their own segments, their own preference, depending on the emotion and the character of the listener. At 1989, Robert E. Thayer proposed a two dimensional model of the mood and emotion (of the music) through his research. The aforementioned model is hoped to be able to used as a reference in terms of sorting music based by their mood. The objective of this research is the classification of music samples by using their audio features through the usage of Fuzzy KNN method. There will be four mood categories used at this research, namely Anger, Happy, Calm and Sad. Extraction of “energy”, ‘tonality” and “tempo” will be done using the library of Matlab MIRToolbox. Aforementioned “MIRToolbox” is a library collection, excelled in the processing of unique attributes in music, in the perspective of Music Information Retrieval. Generally the classification results of audio features used in the aforementioned software, is satisfactory, as long as the K parameter is correct. Fuzzy system is added in order to count the membership degree of the samples.