PENERAPAN LEARNING VECTOR QUANTIZATION (LVQ) UNTUK KLASIFIKASI STATUS GIZI ANAK

• المؤلفون الرئيسيون: , ELVIA BUDIANITA, , Drs. Widodo Prijodiprojo ,M.Sc.,EE
• التنسيق: Theses and Dissertations NonPeerReviewed
• منشور في: [Yogyakarta] : Universitas Gadjah Mada 2013
• الموضوعات: ETD
• الوصول للمادة أونلاين: https://repository.ugm.ac.id/119369/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=59366
• المؤسسة: Universitas Gadjah Mada

One of the indicators of nutritional status that has been tested in a variety of nutrition program and activities is anthropometry. The classification of children nutrient status that commonly used is based on body weight for age index by using zscore table list or deviation standard WHO NCHS (National Centre for Health Statistic). In this study, variable that used in this appraisal are genre, weight, high, infection disease, appetite, and father�s work. The data that used is health recapitulation of student in Sekolah Dasar Batupanjang, Rupat Subdistrict, Bengkalis Regency, Riau Province in 2012. Neural network method that used in this classification are Learning Vector Quantization (LVQ) and one of it develop algorithm, it is LVQ3. LVQ method is a pattern classification that each output unit represents a category or class. The process that happen in each neuron is calculate the nearest distance between a input vector to relevant integrity. Based on result of the study and discussion, LVQ3 algorithm is better applied for children nutrient status classification than LVQ1. Parameter value of learning rate (α) = 0.05, value of minimal learning rate (Mina) = 0.02, value of subtracter α = 0.1, and value of window (ε) = 0.2 that use in LVQ3, is a good parameter score that effective and efficient in appraising classification of nutrient status for elementary student because it has appropriate with all of target (100%). Using of window parameter (ε) in LVQ3 neural network effect positive impact, that is can increase the perform in classification than without using window (LVQ1).