Research On Method Of Cancer Diagnosis Based On Machine Learning

The regular medical diagnosis process can be discribed as a decision-making process that the doctor judges the etiology according to their own clinical experience and the patient's symptoms.Early diagnosis and treatment of cancer and other complex diseases can greatly improve the user's cure rate,and the wrong diagnosis may cause the patient to die.This kind of decision-making problem mainly depends on the doctor's practical work experience while the experience come from long-term pracice and study,and the diagnosis has received a greater impact on subjective factors and the external environment.At present,the rearch on expert system using mathematical and reasoning model is more and more popular,and these expert system is used to help the doctor to make better choice.This type of medical expert system uses the physician's experience and diagnostic knowledge to design judgment rules and builds a knowledge base that uses reasoning to diagnose.However,these rule-based expert systems have its own disadvantages such as reasoning inefficiency and the difficulties of acquisiting and discribe the knonlege of medical diagnosis,thus the effect on helping diagnosis of complex ilnesses is limited.With the development of machine learning and big data,a large number of medical data provides a prerequisite for the use of machine learning algorithms to design diagnostic assistance decision systems.In this paper,breast cancer is chosen as experimental case,and rearch on how to design assistant system to help diagnosis using machine learning algorithm.First,this article describes the main means of intelligent medical care and the application of machine learning algorithm in intelligent medical care.Scondly,we introduced the theoretical basis and their application in intelligent diagnosis of support vector machine,genetic algorithm and particle swarm optimization algorithm in detail.The genetic algorithm and particle swarm optimization algorithm are combined with support vector machine to improve the performance of support vector machines,named as GA-SVM and PSO-SVM.Finally,experiments were conducted using Wisconsin Breast Cancer Data set to verify the feasibility of the model.Based on the experiment results of GA-SVM and PSO-SVM,we propose a method of cancer diagnosis based on integrated learning.The experiment result shows that The method of integrated learning has a good effect on breast cancer cases.The method of machine learning provides a new method for the diagnosis of breast cancer and with a very important practical significance and application value.