Wearable Sensors in Health Monitoring Systems S.Sivasakthi A.Rajeswari Assistant Professor, Department of Computer Science, Assistant Professor, Department of Computer Science, G.Venkataswamy Naidu College (SFC), Kovilpatti. G.Venkataswamy Naidu College (SFC), Kovilpatti. sakthiravi04@gmail.com rajee.alwar@gmail.com Abstract Recent years have perceived an increase in the progress of wearable sensors for health monitoring systems. This increase has been due to several issues such as development in sensor technology as well as focused efforts on political and investor levels to promote projects which address the need for providing new methods for care given increasing challenges with an aging population. In this system is about study of how the data is treated and processed. This paper provides latest methods and algorithms used to analyze data from wearable sensors used for physiological monitoring of vital symbols in healthcare services. This paper outlines the data mining tasks that have been applied such as prediction, anomaly detection and decision making when considering in particular continuous time series measurements and detailed about the suitability of particular data mining and machine learning methods used to process the physiological data and provides an overview of the properties of the data sets used in experimental support. Keywords: Introduction ;Data mining for wearable sensors; Data mining Approach; ; Machine learning technique; Datasets and its properties; 1. Introduction With the increase of healthcare services to i mplement the health monitoring system continuously without hospitalization using wearable sensors, the need to mine and process the physiological measurements is growing significantly. Advances in data mining for health monitoring systems have led to provide proactive information However, as the field progresses and more works consider utilization in real settings, data mining techniques that consider the specific challenges which develop from data coming from wearable sensors is of ever growing significance. It includes not only traditional pattern recognition and anomaly detection but also must consider decision based systems which can handle context awareness, and subject specific models and personalization. Data mining reviews for healthcare and sensors most of them are related to general studies for healthcare i.e., well known problems in healthcare with simple and routine data mining approaches categorized the main challenges of sensor data mining in five following stages: acquisition, preprocessing, transformation, modelling and evaluation. Data mining algorithms mainly classified in two categories (1) Descriptive or unsupervised learning ( i.e., clustering, association, and summarization) and (2) Predictive or supervised learning ( i.e., classification, regression). However, they are needing deeper insight into the appropriateness of the algorithms for handling the special characteristics of the sensor data in health monitoring systems. 2. Data Mining for Wearable Sensors Research area of health monitoring systems has moved from simple reasoning of wearable sensor readings (like calculating the sleep hours) to the higher level of data processing in order to give more information that is valued to the end users, i.e. healthcare services have been converging on deeper data mining tasks to have deeper knowledge discovery. Three types of data mining tasks are predominant. They are: i) Prediction, ii) Anomaly detection iii) Diagnosis In Figure – 1 the first dimension involves in which the monitoring occurs. The most monitoring applications which consider home settings or remote monitoring deal primarily with prediction and anomaly detection whereas the applications in clinical settings are typically focused on diagnosis. This fact is explained by the growing need to have a more precautionary approach (prediction) via wearable sensors and to consider the chance to enable living in home environments by increasing the sense of security (alarm). Similarly, in clinical settings information is available in order to provide diagnosis and assist in decision making. A second dimension in the Figure shows the main data mining tasks in wearable sensors with respect to the type of issues used. For patients with known medical records, both diagnosis and specifically the possibility to raise alarms are key tasks. For health monitoring which include healthy individuals who want to ensure the maintenance of good health, prediction and anomaly detection are used .The final dimension depicted in the Figure considers the three main data mining tasks in relation to how the data is processed. For all three tasks data has been addressed both in an online and offline manner. International Journal of Advanced Research in Basic Engineering Sciences and Technology (IJARBEST) ISSN(Online) : 2456-5717 108 Vol. 3, Special Issue 36, March 2017