Guerrero Alonso, Juan IgnacioParejo Matos, AntonioPersonal Vázquez, EnriqueMonedero Goicoechea, Iñigo LuisBiscarri Triviño, FélixBiscarri Triviño, JesúsLeón de Mora, Carlos2021-10-202021-10-202017Guerrero Alonso, J.I., Parejo Matos, A., Personal Vázquez, E., Monedero Goicoechea, I.L., Biscarri Triviño, F., Biscarri Triviño, J. y León de Mora, C. (2017). From Rule Based Expert System to High-Performance Data Analysis for Reduction of Non-Technical Losses on Power Grids. International Journal on Advances in Intelligent Systems, 10 (1-2), 136-146.1942-2679https://hdl.handle.net/11441/126706The Non-Technical Losses represent the non-billed energy due to faults or illegal manipulations in customer facilities. The objective of the Midas project is the detection of Non-Technical Losses through the application of computational intelligence over the information stored in utility company databases. This project has several research lines, e.g., pattern recognition, expert systems, big data and High Performance Computing. This paper proposes a module which uses statistical techniques to make patterns of correct consumption. The main contribution of this module is the detection of cases, which are usually classified as consumers with Non-Technical Loss increasing the false positives and decreasing the total success rate. This module is integrated with a rule based expert system made up of other modules, such a text mining module and a data warehousing module. The correct consumption patterns (consumers without Non-Technical Losses) are generated using rules, which will be used by a rule based expert system. Two implementations are proposed. Both of them provided an Intelligent Information System to reach unapproachable goals for inspectors. Additionally, some highlighted cases of detected patterns are described.application/pdf11engAttribution-NonCommercial-NoDerivatives 4.0 Internacionalhttp://creativecommons.org/licenses/by-nc-nd/4.0/Non-technical lossesPattern recognitionExpert systemBig Data analyticsHigh performance computingHigh performance data analysisinfo:eu-repo/semantics/articleinfo:eu-repo/semantics/openAccess