Explainable History Distillation by Marked Temporal Point Process

Abstract

Explainability of machine learning models is mandatory when researchers introduce these commonly believed black boxes to real-world tasks, especially high-stakes ones. In this paper, we build a machine learning system to automatically generate explanations of happened events from history by ca based on the tpp. Specifically, we propose a new task called ehd. This task requires a model to distill as few events as possible from observed history. The target is that the event distribution conditioned on left events predicts the observed future noticeably worse. We then regard distilled events as the explanation for the future. To efficiently solve ehd, we rewrite the task into a 01ip and directly estimate the solution to the program by a model called model. This work fills the gap between our task and existing works, which only spot the difference between factual and counterfactual worlds after applying a predefined modification to the environment. Experiment results on Retweet and StackOverflow datasets prove that model significantly outperforms other ehd baselines and can reveal the rationale underpinning real-world processes.

Tasks

Reproductions