Previous activity-based studies have shown that behavioural outcomes are the result of complex and multidimensional processes. In this context, identifying and characterizing discrete mobility profiles through the classification of people’s behavior is particularly attractive. By facilitating the interpretation of complex, multidimensional processes, such an exercise could help to efficiently target transport policy decisions. The purpose of this paper is to identify mobility strategy profiles considering four key dimensions: time, space, money, and social interaction. Data from a seven-day activity, travel, expenditure, and social interaction diary applied to a sample of residents from the city of Concepcion, Chile, is used. A two-step approach based on machine learning techniques is adopted. First, we use a Self-Organizing Map algorithm to identify seven distinct mobility strategies, each characterized by distinctive behaviors. These profiles are identified through 18 weekday and weekend daily behavior variables measuring the four key dimensions mentioned above, with an explicit focus on transport mode use. In the second step, we use a Decision Tree algorithm to profile the mobility strategies by means of personal and household sociodemographic variables. The results show interesting links among the dimensions of analysis within these profiles, such as connections between monetary expenditure on leisure and daily social interaction, and that profiles with higher private vehicle modal split tend to present higher levels of social interaction with people in their social network than public transit users.

Time, space, money, and social interaction: Using machine learning to classify people’s mobility strategies through four key dimensions

We propose weighted repeated median filters and smoothers for robust non-parametric regression in general and for robust signal extraction from time series in particular. The proposed methods allow to remove outlying sequences and to preserve discontinuities (shifts) in the underlying regression function (the signal) in the presence of local linear trends. Suitable weighting of the observations according to their distances in the design space reduces the bias arising from non-linearities. It also allows to improve the efficiency of (unweighted) repeated median filters using larger bandwidths, keeping their properties for distinguishing between outlier sequences and long-term shifts. Robust smoothers based on weighted L1- regression are included for the reason of comparison.

Weighted Repeated Median Smoothing and Filtering

Cities have a common characteristic in the form of land utilisation, which is dominated by built-up areas. Tourism is an essential aspect of city development because it can involve the identity of the city. Historical buildings, landmarks, shopping centres and museums are generally interesting places for tourists to visit. Yogyakarta, the research area, is synonymous as a city of culture and of students. Knowledge of the spatial clustering patterns of tourists can be one of the references for urban development. Social media data were used in the study as an alternative to direct data collection, which requires considerable resources. Flickr and Twitter were used as proxies to dete rmine the distribution of tourists, and the DBSCAN and HDBSCAN clustering algorithms were used to determine the centres of tourist activity. Furthermore, Flickr data were analysed temporally to determine the impact of the COVID-19 pandemic on tourism in Yogyakarta City. The clustering of social media data results shows that there are several city hotspots, besides the already well-known tourist attractions. Apart from city landmarks, several other tourist hotspots were revealed through the clustering process, such as accommodation, shopping centres, entertainment venues and souvenir shops, which also support tourism activities. The impact of COVID-19 on tourism in Yogyakarta City can be reflected through the number of uploaded photos by tourists on Flickr, which has decreased since March 2020.

/pdf/revealing-tourist-hotspots-in-yogyakarta-city-based-on-4dyloeqsqe.pdf

Revealing tourist hotspots in yogyakarta city based on social media data clustering

Nowadays, precise data of movements of public transport can be collected. Specifically, for each bus, geoposition can be regularly obtained and stored. In this context, an attempt to build a model to represent behavior of busses, and predict their delays, is discussed. 

Machine Learning for Bus Travel Prediction

Modeling and prediction of traffic systems is a challenging task due to the complex interactions within the system. Identification of significant regressors and using them to improve travel time predictions is a concept of interest. In previous studies, such regressors were identified offline and were static in nature. In this study, an iterative joint clustering and prediction approach is proposed to accurately predict spatiotemporal patterns in travel time. The clustering module is tied to the prediction module, and a prediction model is trained on each cluster. The combined clustering and prediction are then iterated until a chosen metric is optimized. This orients clusters of data towards prediction while enabling model development on subsets of travel time data with similar prediction complexity. The clusters created using the joint clustering and prediction approach confirmed to the real-world traffic scenario, forming clusters of high travel time at busy intersections and bus stops across the study stretch and forming clusters of low travel time in the sub-urban areas of the city. Further, a comparison of the developed framework with base methods demonstrated a decrease in prediction errors by at least 22.83%. This indicates that creating clusters of data that are sensitive to the quality of predictions using the joint clustering and prediction framework improves the accuracy of travel time predictions. The study also proposes criteria for choosing the best predictions when cluster-based predictions are used.

Joint clustering and prediction approach for travel time prediction

Providing accurate and reliable travel time information to travellers is essential to improve the quality of public transit systems. With the availability of the latest technologies, it has become ...

Evaluation of Clustering Algorithms for the Prediction of Trends in Bus Travel Time

Studying patterns in traffic data is a basic analysis to understand the system. In this study, a large amount of bus travel data collected using vehicle tracking devices is analyzed for patterns. Travel time, in general, follows both spatial and temporal patterns. Spatial patterns are expected because travel times in particular sections on a roadway can be following similar patterns. For example, sections with a bus stop in it may show similar patterns due to stopping at the bus stops. The present study explores the use of data-driven approaches, primarily clustering, to identify the spatial patterns in bus travel times. Discrete Wavelet Transform (DWT) is used to extract trends from the travel time measurements. Two popular clustering algorithms - k-means and hierarchical clustering algorithms are used in this study to identify the spatial patterns and group sections with similar characteristics. Once the spatial patterns are obtained, the historic database is searched to identify similar cluster patterns and travel time trends are predicted using Pattern Sequence-based Forecasting (PSF) algorithm. The performance of the proposed algorithm for the prediction of travel time trends of trips occurring during peak and off-peak hours of a day was then compared based on prediction errors.

Prediction Of Trends In Bus Travel Time Using Spatial Patterns

The prediction of bus travel time with accuracy is a significant step toward improving the quality of public transportation. Drawing meaningful inferences from the data and using these to aid in prediction tasks is always an area of interest. Earlier studies predicted bus travel times by identifying significant regressors, which were identified based on chronological factors. However, travel time patterns may vary depending on time and location. A related question is whether the prediction accuracy can be improved with the choice of input variables. The present study analyzes this question systematically by presenting the input data in different ways to the prediction algorithm. The prediction accuracy increased when the dataset was grouped, and separate models were trained on them, the highest accurate case being the one where the data-derived clusters were considered. This demonstrates that understanding patterns and groups within the dataset helps in improving prediction accuracy.

/pdf/effects-of-data-characteristics-on-bus-travel-time-27d1bghb.pdf

Effects of Data Characteristics on Bus Travel Time Prediction: A Systematic Study

Improving the accuracy of travel time predictions depends on providing the correct inputs as well as the prediction algorithm used. Clustering algorithms can be used to identify the patterns in the data, which can improve the inputs to the prediction algorithm. The feature vectors used for clustering greatly affect the clusters formed and, ultimately, the prediction performance. Clustering being an unsupervised learning technique, the accuracy or correctness of the cluster formed can not be evaluated directly. A possible solution for this would be to link the problem with prediction accuracy and choose the feature vector combination with maximum prediction accuracy. The present study analyses the use of different feature vectors for clustering and the effect on travel time predictions. Here, three cases, namely, travel time alone, travel time along with features such as time of the day, section index, and day of the week as numerical features and as a mix of categorical and numerical feature vectors, are studied. The effects of using each of these cases as clustering feature vectors on travel time predictions are evaluated. It is observed that the prediction accuracy is the highest when only travel times are used as a clustering feature vector. The study demonstrates the importance of choosing the correct feature vectors for clustering and its effect on a final application, namely, travel time prediction.

Hima Elsa Shaji

Papers

Evaluation of Clustering Algorithms for the Prediction of Trends in Bus Travel Time

Prediction Of Trends In Bus Travel Time Using Spatial Patterns

Joint clustering and prediction approach for travel time prediction

Effects of Data Characteristics on Bus Travel Time Prediction: A Systematic Study

Effects of Clustering Feature Vectors on Bus Travel Time Prediction: A Case Study