KLASIFIKASI STROKE MENGGUNAKAN METODE LEARNING VECTOR QUANTIZATION (LVQ)

Stroke is a brain attack that arises suddenly where there is a partial or complete disruption of brain function as a result of a disruption in blood flow due to a blockage or rupture of certain blood vessels in the brain. This research uses the Learning Vector Quantization (LVQ) method. LVQ is a method for learning in supervised competitive layers. A competitive layer will automatically learn to classify input vectors. This study uses 7 initial symptoms namely awareness (X1), nauseous vomit (X2), headache (X3), difficulty speaking (X4), limited movement (X5), weakness (X6) and seizures (X7). The results of this study are getting a MATLAB Graphic User Interface (GUI) program as a decision support tool in detecting stroke and classifying stroke by using a learning rate (?) of 0.1 and a declining rate (dec ?) of 0.75 with the number of epochs was 5 iterations, so that the accuracy of the classification of stroke in Central Sulawesi using the LVQ method was 96.08%.

keywords: Learning Vector Quantization, Stroke, Classification, Learning Rate