Research And Implementation Of The Data Mining Algorithms Of The Internet Of Things In Healthcare Based On Hadoop

The development level of medical and health system is directly related to people’s health and the realization of the Chinese dream, is also a hot topic of the whole society. In the key period of the healthcare reform, combining with the technology of the Internet of things and cloud computing, the society should strengthen the level of informatization in the field of healthcare. The Internet of things in healthcare should realize accurate and real-time perception of medical information through the technology of advanced sensing, achieve interconnectivity in the field of healthcare through the convenient comprehensive communication technology, and can analysis and forecast health data through the technology of efficient data processing.The data processing flow of the Internet of things in healthcare is mainly receiving health data which intelligent data acquisition terminal sends, storing health information, filtering data and mining data in a distributed form to make medical experts analyze the condition more efficient. To analyze and mine data of the Internet of things in healthcare in Hadoop, the paper researches and analyzes the source code of Hadoop, designs algorithms of distributed data filtering to filter redundant data, researches several data mining algorithms, and verifies the results right. These algorithms and models provide nice forecasting functions for healthy service.When mining the health data, first the paper derives the idea of ant colony, designs and realizes DKBAC clustering algorithm in distributed form on Hadoop. Through experiment DKBAC compares with the relevant clustering algorithms on the accuracy, the average recall rate and time for the human physiological data. To meet the demand of analysis in the Internet of things in healthcare, it needs to study and achieve more algorithms and models. The paper researches and designs random forest and FP-Growth in a distributed form on Hadoop, which compare with relevant algorithms in performance for physiological data. After researching these distributed algorithms of data mining, the paper summarizes the general principles and the applicable scope of designing distributed algorithms, points out the direction for researching more distributed algorithms in the Internet of things in healthcare.