2020/01/21 by Daniele Gammelli, Gammelli, Daniele, Inon Peled +9 · 2 citations
Engineering · Social Sciences · #FOS: Computer and information sciences #Human Mobility and Location-Based Analysis #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Traffic Prediction and Management Techniques #Transportation and Mobility Innovations
paper · pdf · doi:10.48550/arxiv.2001.07402
openalex publication_date 2020/01/21 · openalex created_date 2022/07/24 · openalex updated_date 2026/07/28
Transport demand is highly dependent on supply, especially for shared\ntransport services where availability is often limited. As observed demand\ncannot be higher than available supply, historical transport data typically\nrepresents a biased, or censored, version of the true underlying demand\npattern. Without explicitly accounting for this inherent distinction,\npredictive models of demand would necessarily represent a biased version of\ntrue demand, thus less effectively predicting the needs of service users. To\ncounter this problem, we propose a general method for censorship-aware demand\nmodeling, for which we devise a censored likelihood function. We apply this\nmethod to the task of shared mobility demand prediction by incorporating the\ncensored likelihood within a Gaussian Process model, which can flexibly\napproximate arbitrary functional forms. Experiments on artificial and\nreal-world datasets show how taking into account the limiting effect of supply\non demand is essential in the process of obtaining an unbiased predictive model\nof user demand behavior.\n