Robust Searching-based Gradient Collaborative Management in Intelligent Transportation System

With the rapid development of big data and Internet Things(IoT), traffic from an Intelligent Transportation System(ITS) is becoming more accessible. To understand simulate patterns data, Multimedia Cognitive Computing(MCC) efficient practical approach. Distributed Machine Learning(DML) has been trend to provide sufficient computing resources efficiency for MCC tasks handle massive complex models. DML can speed up computation with those but introduces communication overhead. Gradient collaborative management or gradient aggregation in a critical task. An managing algorithm schedules ITS improve performance tasks. However, existing typically rely on specific physical connection matrices, which have low robustness when malfunction occurs. In this paper, we propose Robust Searching-based Collaborative Management(RSGCM) System, ring-based across devices deal malfunction. RSGCM provides solutions various kinds matrices acceptable amount training time. Our experimental results shown that varieties than state-of-the-art schedules. also increases since it restore system’s functionality time device breakdown happens.