From Statistics to Analytics: A Novel Quantum-Based Framework for Traffic Analyses

In the evolving landscape of advancements cities are developing rapidly. To meet basic needs people are increasingly using personal vehicles, which increases traffic and the complexity of driving. The primary goal of this research is to ingest, store, and analyze the traffic dataset using a novel technique. While surveying the existing methods it was found that data ingestion and management techniques were quite limited; the existing studies primarily focused on analyzing data using deep network architectures, which require higher computational resources due to their complex architecture. To overcome these gaps we proposed the incremental histogram and quantile partition as quantum approximate optimization algorithm (IHQ-QAOA) technique, which utilizes histogram binning and quantile techniques to ingest data more efficiently with minimal resources. The QAOA technique addresses the clusters by optimizing the cost and identifying the lane change behavior of the vehicles. We extracted insights to analyze traffic patterns using online analytical processing techniques (pivoting, drill-through, slicing, and dicing), which supported the segregation of features from section to lane level, resulting in better data visualization. The proposed techniques are compared with the state-of-the-art using the Next Generation SIMulation (NGSIM) and New Era of Urban Traffic Monitoring with Aerial footage (pNEUMA) datasets, resulting in an average data ingestion efficiency of 75.53%; for lane change analysis, the pNEUMA dataset was tested on a model trained on the NGSIM dataset, and overall accuracies of more than 0.96 were obtained. These achievements show that the suggested approach is reliable and effective for traffic data analysis.