An open question regarding the possibility of completely automating the approach presented in Chap. 4 is addressed in this last methodological chapter. It has already been discussed in the previous chapters how the proposed BENEDICT approach still lacks an autonomous process for the generation of distal information in order to be completely automated. Indeed, the procedure for creating the causal explanations does not require any other updates given the proposed methods applied to the RL scenarios.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Auto-BENEDICT: Combining Importance Metrics and BENEDICT for an Automatic Explanation Generator

  • Rudy Milani

摘要

An open question regarding the possibility of completely automating the approach presented in Chap. 4 is addressed in this last methodological chapter. It has already been discussed in the previous chapters how the proposed BENEDICT approach still lacks an autonomous process for the generation of distal information in order to be completely automated. Indeed, the procedure for creating the causal explanations does not require any other updates given the proposed methods applied to the RL scenarios.