Learning Marked Temporal Point Process Explanations based on Counterfactual and Factual Reasoning
This work addresses trustworthiness concerns in high-stakes applications using MTPP models, offering an incremental improvement in explanation methods for event sequence predictions.
This study tackled the problem of explaining predictions from neural network-based Marked Temporal Point Process (MTPP) models by identifying minimal subsets of historical events that maintain prediction accuracy, proposing a method combining counterfactual and factual reasoning to avoid irrational explanations. Experiments showed the proposed CFF method outperformed baselines in explanation quality and processing efficiency.
Neural network-based Marked Temporal Point Process (MTPP) models have been widely adopted to model event sequences in high-stakes applications, raising concerns about the trustworthiness of outputs from these models. This study focuses on Explanation for MTPP, aiming to identify the minimal and rational explanation, that is, the minimum subset of events in history, based on which the prediction accuracy of MTPP matches that based on full history to a great extent and better than that based on the complement of the subset. This study finds that directly defining Explanation for MTPP as counterfactual explanation or factual explanation can result in irrational explanations. To address this issue, we define Explanation for MTPP as a combination of counterfactual explanation and factual explanation. This study proposes Counterfactual and Factual Explainer for MTPP (CFF) to solve Explanation for MTPP with a series of deliberately designed techniques. Experiments demonstrate the correctness and superiority of CFF over baselines regarding explanation quality and processing efficiency.