Using Machine Learning on Sensor Data

Abstract

Extracting useful information from raw sensor data requires specific methods and algorithms. We describe a vertical system integration of a sensor node and a toolkit of machine learning algorithms for predicting the number of persons located in a closed space. The dataset used as input for the learning algorithms is composed from automatically collected sensor data and additional manually introduced data. We analyze the dataset and evaluate the performance of two types of machine learning algorithms on this dataset: classification and regression. With our system settings, the experiments show that augmenting sensor data with proper information can improve prediction results and also the classification algorithm performed better.