Large scale data analysis on the Monday syndrome problem in traffic flow

Within the field of traffic data analysis, the aggregation of week to week trafficdata is performed regularly. Due to the known effect of human behavior on traffic patterns, a previous study on weekdays versus weekends traffic focused on capacity and flow [3]. One of the results of the paper included the statement, ”It is presumed that there is no substantial difference between workday capacities, as the driver population is in general identical on most days.” This presumption was based on the idea that driver population is consistent between weekdays. However, the assumption that drivers between Mondays and Fridays are generally identical may not be truly representative of human behavior. It is believed that there are distinct traffic patterns between Mondays andFridays compared to the rest of the working week. This hypothesis comes from decisions to choose certain days of the week to take traffic data. For example, a study on traffic speed prediction using k-nearest neighbor took traffic data from Mondays through Thursdays for each week [5]. To further investigate the hyposthesis proposed, large scale traffic data provided by the Caltrans Performance Measurement System database [1] will be used to study week to week traffic data focusing on the weekdays.M.S.Includes bibliographical reference