Terrorist Organization Behavior Prediction Algorithm Based on Context Subspace

摘要: Researchers have developed methods to predict a terrorist organization's probable actions (such as bombings or kidnappings). The can be affected by the context of organization. Thus, variables used improve accuracy forecasting behavior. Those algorithms based on similarity suffer serious drawback that it result in algorithm's fluctuation and reduce prediction if not all attributes are detected. A algorithm PBCS (Prediction Based Context Subspace) subspace is proposed this paper. first extracts according association between behavior attributes. Then, predicts extracted subspace. uses improved spectral clustering method obtain It concerns distribution data samples, label information local process extracting Experimental results artificial dataset MAROB show paper only but also fluctuation.