Mendiagnosa Penyakit Mata Menggunakan Jaringan Saraf Tiruan dengan Menggunakan Metode Backpropagation dan Hopfield

Abstract

The eye is the most important organ of living things, especially humans. The main function of the eye is for the sense of sight. The eye has important parts such as the cornea, pupil, retina, sclera, eye lens and others. Eye health is an important thing for the health of the human body, because the eye is very helpful for carrying out daily activities. Everyday. Almost every human activity uses the eyes, for example reading, working, watching television, writing, driving, etc. so that many people agree that the eyes are the five most important senses. If the eye has eye disorders or diseases, it will have fatal consequences for human life. The problem faced in designing this system is the problem of diagnosing those affected by the disease. The neural network methods used are backpropagation and Hopfield to conclude which method has Hopfield to conclude which method has better accuracy in identifying which method has Hopfield. object discussed. In the backpropagation method, the pattern is trained through the first three phases, namely the forward propagation phase, the back propagation phase, and the weight change phase until the stopping conditions are met, while in the Hopfield method the training is carried out by doing a dot product between the input pattern vector and the vector. The Hopfield network is said to reach a maximum value if a stable pattern is recalled. Based on the results of trials on eye disease, it is known that the Hopfield method can recognize patterns faster than the backpropagation method with an average recognition time of 2.46 and 5.67 seconds.