Kühlwein, Daniel; Schulz, Stephan; Urban, Josef E-MaLeS 1.1. (English) Zbl 1381.68273 Bonacina, Maria Paola (ed.), Automated deduction – CADE-24. 24th international conference on automated deduction, Lake Placid, NY, USA, June 9–14, 2013. Proceedings. Berlin: Springer (ISBN 978-3-642-38573-5/pbk). Lecture Notes in Computer Science 7898. Lecture Notes in Artificial Intelligence, 407-413 (2013). Summary: Picking the right search strategy is important for the success of automatic theorem provers. E-MaLeS is a meta-system that uses machine learning and strategy scheduling to optimize the performance of the first-order theorem prover E. E-MaLeS applies a kernel-based learning method to predict the run-time of a strategy on a given problem and dynamically constructs a schedule of multiple promising strategies that are tried in sequence on the problem. This approach has significantly improved the performance of E 1.6, resulting in the second place of E-MaLeS 1.1 in the FOF divisions of CASC-J6 and CASC\(\@\)Turing.For the entire collection see [Zbl 1264.68002]. Cited in 4 Documents MSC: 68T15 Theorem proving (deduction, resolution, etc.) (MSC2010) Software:E-MaLeS; MiniSat; MaLARea; MPTP 0.2; E-SETHEO; Gandalf; z3; SATzilla; VAMPIRE; TPTP; Isabelle/HOL PDFBibTeX XMLCite \textit{D. Kühlwein} et al., Lect. Notes Comput. Sci. 7898, 407--413 (2013; Zbl 1381.68273) Full Text: DOI