APREP-DM: a Framework for Automating the Pre-Processing of a Sensor Data Analysis based on CRISP-DM
Hiroko Nagashima, Yuka Kato · 2019
The need for analyzing data is increasing at an unprecedented rate. Well-known examples include customer behavioral patterns in shops, the autonomous motion of robots, and fault prediction. Pre-processing of data is essential for achieving accurate results. This includes detecting outliers, handling missing data, and data formatting, integration, and normalization. Pre-processing is necessary for eliminating ambiguities and inconsistencies. We here propose a framework called APREP-DM (for the Automated PRE-Processing for Data Mining) applicable to data analysis, including using sensor data. We evaluate two types of perspectives: (1) considering pre-processing in a test-case scenario involving pedestrian trajectory tracking, and (2) comparing APREP-DM with the outcomes of other existing frameworks from four different perspectives. We conclude that APREP-DM is suitable for analyzing sensor data.