-, Subairi (2018) Klasifikasi Masa Ovulasi Menggunakan Metode Jaringan Saraf Tiruan Propagasi Balik Dan K-Nearest Neighbor. Magister thesis, Universitas Brawijaya.
Abstract
Masa subur pada wanita dewasa adalah periodik dan berhubungan dengan siklus menstruasi, tetapi fakta bahwa puncak kesuburan terjadi pada titik tengah perkiraan antara menstruasi tidak diketahui sampai pada era tahun 1930-an. Untuk mengetahui masa subur telah banyak dikembangkan alat monitor kesuburan dengan berbagai metode antara lain, metode lendir servik, monitor hormon Luteinizing urin, monitor suhu tubuh basal, observasi diameter folikular, monitor folikel hormon antimulerian, observasi online, monitor hormon kesuburan elektronik, metode peak day, dan metode standard day. Selain metode tersebut, monitoring masa subur juga bisa dideteksi dengan memeriksa kristal cairan air liur (ferning saliva). Pada penelitian ini didiskusikan dan dijelaskan hasil penelitian tentang bagaimana mendeteksi dan mengklasifikasikan masa ovulasi pada wanita berdasarkan citra pola ferning saliva menggunakan perbandingan metode Jaringan Saraf Tiruan Propagasi Balik (JST-PB) dan metode K-Nearest Neighbor (K-NN) sehingga dapat mengklasifikasikan masa tidak subur (infertile period), transisi ke masa subur (intermediate period), dan puncak masa subur (fertile period) wanita dengan akurasi yang baik. Data masukan berupa gambar ferning saliva kemudian diolah pada proses pengolahan citra menggunaan ekstraksi fitur tekstur menggunakan metode gray-level co-occurence matrices (GLCM). Terdapat lima parameter fitur yang dipakai yaitu angular second moment (ASM), contrast, inverst different moment (IDM), entropy, dan correlation, yang diambil dari rata-rata hasil masing-masing sudut 0°, 45°, 90°, dan 135°. Sedangkan data keluaran adalah salah satu dari tiga klasifikasi dari masa ovulasi, yaitu infertile, intermediate, dan fertile. Dari hasil penelitian ini didapat bahwa metode JST-PB berhasil melakukan pelatihan dan pengujian klasifikasi citra ferning saliva untuk menentukan masa ovulasi wanita dengan akurasi klasifikasi sebesar 100.0%, 86.7%, dan 86.7%, sedangkan metode K-NN mencapai akurasi sebesar 93.3%, 80.0%, dan 86.7%,, keduanya berurutan untuk klasifikasi infertile, intermediate, dan fertile.
English Abstract
Although the ovulation period of an adult female has a periodic cycle and is related to its menstrual cycle, the fact that the fertility peak occurs on the middle point between menstruations was unknown until 1930’s. Some monitoring devices have been developed for ovulation identification with several methods, e.g. cervix fluid method, Luteinizing urine hormone monitor, basal temperature monitor, follicular diameter observation, antimullerian hormone follicle monitor, online observation, peak day method, and standard day method. Additionally, ovulation detection could be done by checking ferning saliva. This study discusses and explains the results of research on how to detect and classify women's ovulation period based on salivary ferning pattern using comparison of the method of Artificial Neural Network Back-Propagation (ANN-BP) and K-Nearest Neighbor (K-NN) method to classify infertility, transition to fertile periods, and peak fertile women with good accuracy. Input data form ferning images of saliva then processed in image processing uses texture feature extraction using gray-level co-occurrence co-occurence (GLCM) method. There are five feature parameters used: angular second moment (ASM), contrast, inverst different moment (IDM), entropy, and correlation, taken from the average yield of each angle 0 °, 45 °, 90 °, and 135 °. While the output data is one of three classifications of the ovulation period, namely infertile, intermediate, and fertile. The result of this research shows that the method of ANN-BP successfully conducted training and testing of salivary ferning image classification to determine the female ovulation period with accuracy classifications of 100.0%, 86.7%, and 86.7%, while K-NN method reach accuracy equal to 93.3%, 80%, and 86,7%, respectively for infertile, intermediate, and fertile classifications.
| Item Type: | Thesis (Magister) |
|---|---|
| Identification Number: | TES/613.943 402 85/SUB/k/041804669 |
| Uncontrolled Keywords: | OVULATION--DETECTION |
| Divisions: | S2/S3 > Magister Teknik Elektro, Fakultas Teknik |
| Depositing User: | Endang Susworini |
| Date Deposited: | 06 Aug 2019 02:57 |
| Last Modified: | 06 Aug 2019 02:57 |
| URI: | http://repository.ub.ac.id/id/eprint/171008 |
Actions (login required)
 |
View Item |