Optimasi Vektor Bobot Pada Learning Vector Quantization Menggunakan Algoritme Genetika Untuk Identifikasi Jenis Attention Deficit Hyperactivity Disorder Pada Anak

Raissa Arniantya (Fakultas Ilmu Komputer, Universitas Brawijaya)
Budi Darma Setiawan (Fakultas Ilmu Komputer, Universitas Brawijaya)
Putra Pandu Adikara (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 2 No 2 (2018): Februari 2018

One of mental disorder which common happened on children under 7 years old. Child with ADHD characterized by lack of ability to concentrate, excessive behavior, and behavior that spontaneously out of control. Type of ADHD are inattention, hyperactive and impulsive. If child with ADHD unidentified early, it will causes psychosocial problem but not many people are aware about ADHD so they need a system for identify the type of ADHD. System uses classification methods Learning Vector Quantization. Some cases classification, LVQ has weak accuracy so it needs optimization methods Genetic Algorithm (GA) for improve the accuracy. LVQ's weight vector will be optimized by GA through genetic process until generated optimum weight vector which LVQ uses for training and testing process. Testing against LVQ and LVQ-GA generate LVQ's accuracy 77% and LVQ-GA's accuracy 92% with best parameters are population size is 75, crossover rate is 0.6, mutation rate is 0.4, number of generation is 80, learning rate is 0.001, learning rate decrement is 0.1, maximum epoch is 1000 and learning rate minimum is 10-16.One of mental disorder which common happened on children under 7 years old. Child with ADHD characterized by lack of ability to concentrate, excessive behavior, and behavior that spontaneously out of control. Type of ADHD are inattention, hyperactive and impulsive. If child with ADHD unidentified early, it will causes psychosocial problem but not many people are aware about ADHD so they need a system for identify the type of ADHD. System uses classification methods Learning Vector Quantization. Some cases classification, LVQ has weak accuracy so it needs optimization methods Genetic Algorithm (GA) for improve the accuracy. LVQ's weight vector will be optimized by GA through genetic process until generated optimum weight vector which LVQ uses for training and testing process. Testing against LVQ and LVQ-GA generate LVQ's accuracy 77% and LVQ-GA's accuracy 92% with best parameters are population size is 75, crossover rate is 0.6, mutation rate is 0.4, number of generation is 80, learning rate is 0.001, learning rate decrement is 0.1, maximum epoch is 1000 and learning rate minimum is 10-16.One of mental disorder which common happened on children under 7 years old. Child with ADHD characterized by lack of ability to concentrate, excessive behavior, and behavior that spontaneously out of control. Type of ADHD are inattention, hyperactive and impulsive. If child with ADHD unidentified early, it will causes psychosocial problem but not many people are aware about ADHD so they need a system for identify the type of ADHD. System uses classification methods Learning Vector Quantization. Some cases classification, LVQ has weak accuracy so it needs optimization methods Genetic Algorithm (GA) for improve the accuracy. LVQ's weight vector will be optimized by GA through genetic process until generated optimum weight vector which LVQ uses for training and testing process. Testing against LVQ and LVQ-GA generate LVQ's accuracy 77% and LVQ-GA's accuracy 92% with best parameters are population size is 75, crossover rate is 0.6, mutation rate is 0.4, number of generation is 80, learning rate is 0.001, learning rate decrement is 0.1, maximum epoch is 1000 and learning rate minimum is 10-16.One of mental disorder which common happened on children under 7 years old. Child with ADHD characterized by lack of ability to concentrate, excessive behavior, and behavior that spontaneously out of control. Type of ADHD are inattention, hyperactive and impulsive. If child with ADHD unidentified early, it will causes psychosocial problem but not many people are aware about ADHD so they need a system for identify the type of ADHD. System uses classification methods Learning Vector Quantization. Some cases classification, LVQ has weak accuracy so it needs optimization methods Genetic Algorithm (GA) for improve the accuracy. LVQ's weight vector will be optimized by GA through genetic process until generated optimum weight vector which LVQ uses for training and testing process. Testing against LVQ and LVQ-GA generate LVQ's accuracy 77% and LVQ-GA's accuracy 92% with best parameters are population size is 75, crossover rate is 0.6, mutation rate is 0.4, number of generation is 80, learning rate is 0.001, learning rate decrement is 0.1, maximum epoch is 1000 and learning rate minimum is 10-16.