Organ Disorder Identification Through Iris Using Multilayer Perceptron Algorithm

M., Taufiq Tamam and Dian, Nova Kusuma Hardani and Latiful, Hayat (2016) Organ Disorder Identification Through Iris Using Multilayer Perceptron Algorithm. PROCEEDING The 1st InCEAS.

	Text (Cover) Cover InCEAS 2016.pdf Download (1MB)
	Text (Full Paper) Full Paper - Organ Disorder Identification Through Iris.pdf Download (318kB)

Abstract

Human organ condition can be seen through the iris as learned in iridology. Iridology is the study of network structure contains in the iris pattern. By the sign of color, texture, and location of pigment in the iris, the state of someone health can be analyzed. Consider to person’s health, identifying disease as well as potential development is very good topic to research. It also can be used as a highly effective complement to gain physical health and quality of life. It has become an important thing to do research in the identification of organs disorder through the iris pattern. The method used in the identification process is a combination of Independent Component Analysis (ICA) with FastICA and MultiLayer Perceptron algorithm. By mixing three different images, it can be obtained three different outputs with different kurtosis value. From those three outputs, one image with has the highest kurtosis value is considered synonymous with the original image. There are seven statistical characteristic extraction results are used as input in the classification process by the method of MultiLayer Perceptron algorithm which are the average, standard deviation, skewness, kurtosis, energy, entropy, and smoothness. The results of the of classification by MultiLayer Perceptron algorithm produces an accuracy rate of 78.9%, a sensitivity of 86.67% and a specificity of 65.38% for organ disorder identification. As for the normal condition, identification produces 78.9% in accuracy rate, 65.38% in sensitivity and 86.67% in specificity.

Actions (login required)

View Item