A Novel Personal Health Assessment Based On Mobile Telecom Data

With the development of the communication industry and the improvement of people's living standard, the data of mobile telecommunication has deeply reflected people's lifestyle. In order to study the correlation between mobile telecommunication data and personal health status, using the method of data mining to analyze the data of mobile telecommunication is an excellent entry point. Modified extreme learning machine based on extreme learning machine frame will be one of the best and the most innovative choices. The advanced features of the modified extreme learning machine are embodied in its random selection of hidden nodes and intelligent analysis of output weights to reduce the complexity of the algorithm. In theory, the algorithm can give a good result in the output of the general law under the premise of the maximum guarantee speed.This paper uses data mining methods from the perspective of machine learning, using modified extreme learning machine method. The key technology of personal health assessment data in mobile telecommunication is studied.The core work of this study is feature extraction and using machine learning on mobile telecommunication data analysis, the main work of this paper includes:1. The algorithm of data mining is discussed.Three kinds of data mining algorithms involved in this study are deduced ,and the mathematical models of the three algorithm are also deduced. Among the three algorithms, this research uses extreme learning machine (ELM) as frame algorithm of study, and other two algorithms which are indicated as support vector machine (SVM) algorithm and back-propagation algorithm (BP) are used for contrast. Meanwhile,the basic idea of machine learning is theoretically deduced;2. This paper shows the specific research process of the prediction and evaluation of the personal health analysis based on mobile telecom data. This paper describes the pretreatment method of mobile telecom data and the feature extraction process;3. This papershows the modification of extreme learning machine algorithm and shows the simulation results,and the performance of the modified extreme learning machine is given.Meanwhile, this paper shows that the improved algorithm in such scenarios has accurate and efficient data processing ability through simulation experiments on different test data and more complex conditions, and compared with other two algorithms the improved algorithm is an efficient and has lower complexity.