Hidden Markov Model (HMM) Statistical Model and Learning from Data Concepts Context-Aware Framework: A Basic Design in Accelerometer Based Smartphones

Abstract

This paper proposes a novel system focused on the Secret Markov Model (HMM) mathematical model and the analysis of data concepts. The architecture either considers or predicts device specific inferences named 'device states' for potential context-conscious implementations. Context-aware systems involve continuous data collection and analysis from one or more sensor readings. The battery life of the unit must also be increased due to the reality that the continually operating built-in sensors easily deplete the battery of the unit. In this way, the framework is built to satisfy the specifications required for applications and to increase the battery life of the unit. The overarching aim of this chapter is to provide reliable representations of the user condition and to optimize power efficiency. Most notably, the purpose of this research is to establish and explain a generic structure to direct the creation of potential context-aware applications. In addition, issues such as adaptability of the user profile and adaptive sampling are discussed. The proposed framework is tested by models and applied in real-time implementations. According to the findings, the proposed system indicates an improvement in power efficiency of 60 per cent with accuracy varying from 75 per cent to 96 per cent, based on user profiles.