Activity Recognition for a Smartphone Based Travel Survey Based on Cross-User History Data

Abstract

In transport modeling and prediction, trip purposes play an important role since mobility choices (e.g. modes, routes, departure times) are made in order to carry out specific activities. Activity based models, which have been gaining popularity in recent years, are built from a large number of observed trips and their purposes. However, data acquired through traditional interview-based travel surveys lack the accuracy and quantity required by such models. Smartphones and interactive web interfaces have emerged as an attractive alternative to conventional travel surveys. A smartphone-based travel survey, Future Mobility Survey (FMS), was developed and field-tested in Singapore and collected travel data from more than 1000 participants for multiple days. To provide a more intelligent interface, inferring the activities of a user at a certain location is a crucial challenge. This paper presents a learning model that infers the most likely activity associated to a certain visited place. The data collected in FMS contain errors or noise due to various reasons, so a robust approach via ensemble learning is used to improve generalization performance. Our model takes advantage of cross-user historical data as well as user-specific information, including socio-demographics. Our empirical results using FMS data demonstrate that the proposed method contributes significantly to our travel survey application.

Who you are is how you travel: A framework for transportation mode detection using individual and environmental characteristics

“…For this study, data from a bespoke developed geo-enabled smart-phone application was used. The use of smart-phone apps in mobility studies has seen an increased interest from researchers in the past years in the fields of activity and transportation mode detection (Montini et al 2015;Kim et al 2014;Widhalm et al 2012). The main advantages of using such an approach are the ease and cost efficiency of data collection process as well as the relatively high spatial accuracy, as the majority of smart-phones are equipped with GPS receivers and accelerometers.…”

Activity Recognition for a Smartphone and Web-Based Human Mobility Sensing System

“…First, we apply two-fold validation where we keep the chronological order of data with k training days and one test day split, k = 1, 2, 3, 4 for every users. In the experiments, we apply different parameter settings: different resolutions of time slot: [10,20,40,60,90,120] For the random subspaces based decision trees (Random Forest (RF)), a dimension of subspace features is chosen based on square root of the total number of feature variables. For decision tree-based (DT) classifiers including RF and bagging of DT (BagDT), the minimum number of observations per tree leaf is set as 1.…”

Inferring Transportation Mode and Human Activity from Mobile Sensing in Daily Life

“…In regards to human activity recognition, Su et al [32] recently presented a general overview of techniques and challenges in performing activity recognition from the mobile phone sensors. Mainly, the research in activity recognition is often treated as a classification problem (for example: [7,10,12,14,26]) since there are labels associated to certain human activities for training and testing phases. Other problem in this domain is related to adaptability to perform real-time activity recognition on continuous streaming of sensor data.…”