Abstract
In the field of medical regularly handles enormous amounts of data. Handling huge data by conventional methods can affect the results. Algorithms for machine learning can be used to find out facts in medical research, in particular for disease prediction. Expert System with artificial intelligence technology, Data mining technology has also been extensively developed, which has promoted the development of for detection and diagnosis of different diseases. The purpose this study aims to know the detection of brain, heart, kidney etc with the help of different algorithms such as Feedforward Backpropagation, Support Vector Machine, Generalized Regression Neural Network, Radial Basis Function and association rules in data mining. This study introduces the above neural network techniques in detail and experiments on medical data collected from hospital. The experimental results show that the some neural network techniques will detect the disease based on medical data collected from hospital and recommend medicine. Applying neural network techniques and data mining techniques based medical data disease detection has greatly improved the level of medical detections and understandings.
Description
In this research work neural network is used to classify the patient’s data for diagnosing proper disease and based on that it gives medicine prescription. In the experimental work, a total of 300 patient’s data has been collected and initially classified into 11 different preliminary headings. Each preliminary hearing is further divided into minimum 4 and maximum 29 sub headings. The Binary transform of patient’s coding is used as input for neural network techniques, like Feed forward Backpropagation, Generalized Regression Neural Network, Radial Basis Function and Support Vector Machine and data mining techniques. In this experimental process each of these neural network techniques are used to generate the neural network classification results. It is revealed that out of these techniques Radial basis function is more prominent technique that gives 97% accuracy in medical prescriptions.
Links and resources
Tags