Inductive programming is concerned with the automated construction of decl- ative-often functional -recursiveprogramsfromincompletespeci?cationssuch as input/output examples. The inferred program must be correct with respect to the provided examples in a generalizing sense: it should be neither equivalent to it, nor inconsistent. Inductive programming algorithms are guided explicitly or implicitly by a language bias (the class of programs that can be induced) and a search bias (determining which generalized program is constructed ?rst). Induction strategiesare either generate-and-testor example-driven.In genera- and-test approaches, hypotheses about candidate programs are generated in- pendently from the given speci?cations. Program candidates are tested against the given speci?cation and one or more of the best evaluated candidates are - veloped further. In analytical approaches, candidate programs are constructed in an example-driven way. While generate-and-test approaches can - in prin- ple - construct any kind of program, analytical approaches have a more limited scope. On the other hand, e?ciency of induction is much higher in analytical approaches. Inductive programming is still mainly a topic of basic research, exploring how the intellectual ability of humans to infer generalized recursive procedures from incomplete evidence can be captured in the form of synthesis methods. Intended applications are mainly in the domain of programming assistance - either to relieve professional programmers from routine tasks or to enable n- programmers to some limited form of end-user programming. Furthermore, in future, inductiveprogrammingtechniquesmightbe appliedtofurtherareassuch as support inference of lemmata in theorem proving or learning grammar rules

Because of its promise to support human programmers in developing correct and efficient program code and in reasoning about programs, automatic program synthesis has attracted the attention of researchers and professionals since the 1970s.

This book focusses on inductive program synthesis, and especially on the induction of recursive functions; it is organized into three parts on planning, inductive program synthesis, and analogical problem solving and learning. Besides methodological issues in inductive program synthesis, emphasis is placed on its applications to control rule learning for planning. Furthermore, relations to problem solving and learning in cognitive psychology are discussed.

This book constitutes the refereed proceedings of the 43rd German Conference on Artificial Intelligence, KI 2020, held in Bamberg, Germany, in September 2020. The 16 full and 12 short papers presented together with 6 extended abstracts in this volume were carefully reviewed and selected from 62 submissions. As well-established annual conference series KI is dedicated to research on theory and applications across all methods and topic areas of AI research. KI 2020 had a special focus on human-centered AI with highlights on AI and education and explainable machine learning. Due to the Corona pandemic KI 2020 was held as a virtual event.

This work discusses the theoretical abilities of three commonly used classifier learning methods and optimization techniques to cope with characteristics of real-world classification problems, more specifically varying misclassification costs, imbalanced data sets and varying degrees of hardness of class boundaries. From these discussions a universally applicable optimization framework is derived that successfully corrects the error-based inductive bias of classifier learning methods on image data within the domain of medical diagnosis. The framework was designed considering several points for improvement of common optimization techniques, such as the modification of the optimization procedure for inducer-specific parameters, the modification of input data by an arcing algorithm, and the combination of classifiers according to locally-adaptive, cost-sensitive voting schemes. The framework is designed to make the learning process cost-sensitive and to enforce more balanced misclassification costs between classes. Results on the evaluated domain are promising, while further improvements can be expected after some modifications to the framework.