Abstract The problems of artificial intelligence, as well as control and decision-making with incomplete or inaccurate information, cover a wide class of problems of abductive explanation, including tasks in terms of cause–effect. This paper is devoted to the logical formation of hypotheses that explain observed effects. Means of representing knowledge and hypothesizing are proposed. A language is introduced that has the property of substitutability. The properties of the language and calculi introduced on its basis provide a convenient combination of deduction and hypothesizing. Unlike well-known logical methods of abduction, the proposed tools provide derivation of hypotheses (minorants) that are necessary and sufficient for a formal explanation of the observed effect. Based on the hypotheses-minorants, in combination with the basic theory of subject domain, reliable causes of the observed effect are formed or relevant circumstances leading to these causes are found. Moreover, in situations where there is also empirical data, these causes and circumstances can also be formed in plausible versions. Examples from technology and medicine are considered.

中文翻译：

---

观察效应解释问题中的溯因推理

摘要 人工智能问题，以及信息不完整或不准确的控制和决策问题，涵盖了一大类溯因解释问题，包括因果关系方面的任务。本文致力于解释观察到的效应的假设的逻辑形成。提出了表示知识和假设的方法。引入了一种具有可替代性的语言。在其基础上引入的语言和微积分的特性提供了演绎和假设的方便组合。与众所周知的推理逻辑方法不同，所提出的工具提供了对观察到的效果的正式解释所必需和充分的假设（次要）的推导。基于假设 - 少数派，结合学科领域的基本理论，形成观察到的效果的可靠原因或找到导致这些原因的相关情况。此外，在也有经验数据的情况下，这些原因和情况也可以形成合理的版本。考虑了技术和医学方面的例子。