A Study Of Medical Instruments’fault Diagnosis Data Mining With Bayesian Classification Algorithm

These days,medical instruments have been an important part in hospital’s day running,different kinds of medical instruments have been used by doctors.However,failures of medical instruments in hospital damage a lot,also bring great inconvenience to patients.Faced with this difficulty,text automatic categorization technology has been introduced in to medical instruments area to realiza the auomatic text classification.This thesis starts with characteristics of short text,short text classification,and its’prospect,then comes focus on——short text of medical instruments failure,and classify this kind of text based on the failure categories that are already achieved.There are so many kinds of text categorization algorithms,among these,we choose Naive Bayes.Cross the realization of short test categorization,we do four kinds of jobs:data collection,short text pre-processing,feature selection and extraction,the realization of Naive Bayes algorithm.After these,we check the generated ratio and precision ratio on short text classifier for evaluation.This is a big step for data mining technology in field of medical instruments.This paper also do this kind of job,that is using GRI relevance algorithm to analyse the relationship between different fault categories,which is the innowation of this paper,by mining and analyzing sup parameter and cong parameter,we conclude the relationship between different fault categories,and hope that the results would provide a set of reasonable early-warning mechanism.