Abstract

With the development of technology, dependence on electricity is increasing day by day. Cost increases as electricity consumption increases. For this reason, individuals aim to save by searching for the optimum benefit-cost relationship. In this study, firstly 1500 rows of data created with day, peak, night values and Roc analysis was used to select the most appropriate algorithm for the data set.Later total electricity consumption in households was analyzed by selecting logistic regression which is data mining algorithm in RapidMiner. The most appropriate tariff selection was provided to the subscribers. As a result, 100% accurate estimation was obtained.

References

[1] Haliloğlu EY, Tutu BE. Short-term electicity power demand forecasting for Turkey. Journalof Yasar University 2018;13(51):243-255.
[2] Nişancı M. Türkiye’de elektrik enerjisi talebi ve elektrik tüketimi ile ekonomik büyümearasındaki ilişki. Sosyal Ekonomik Araştırmalar Dergisi 2005;5(9):107-121.
[3] Akan Y, Tak S. Türkiye elektrik enerjisi ekonometrik talep analizi. Atatürk Üniversitesiİktisadi ve İdari Bilimler Dergisi 2003;17:1-2.
[4] Güloğlu B, Akın E. Türkiye’de hane halkları elektrik talebinin belirleyicileri: sıralı logityaklaşımı. Siyaset, Ekonomi ve Yönetim Araştırmaları Dergisi 2014;2(3):1-20.
[5] Moreno M, Ubeda B, Skarmeta A, Zamora M. How can we tackle energy efficiency in iotbasedsmart buildings?. Sensors;14(6):9582-9614.
[6] Terroso-Saenz F, Gonzalez-Vidal A, Ramallo-Gonzalez AP, Skarmeta AF. An open IoTplatform for the management and analysis of energy data. Future Generation ComputerSystems 2019;92:1066-1079
[7] Chen W, Zhou K, Yang S, Wu C. Data quality of electricity consumption data in a smart gridenvironment. Renewable and Sustainable Energy Reviews 2017;75:98-105.
[8] Gajowniczek K, Zabkowski T. Data mining techniques for detecting householdcharacteristics based on smart meter data. Energies 2015;8(7):7407-7427.
[9] Beckel C, Sadamori L, Staake T, Santini S. Revealing household characteristics from smartmeter data. Energy 2014;78:397-410.
[10] Kavaklioglu K, Ceylan H, Ozturk HK, Canyurt OE. Modeling and prediction of Turkey’selectricity consumption using artificial neural networks. Energy Conversion andManagement 2009;50(11):2719-2727.
[11] Ürük E. İstatistiksel uygulamalarda lojistik regresyon analizi. Marmara Üniversitesi SosyalBilimler Enstitüsü. İstanbul: Yüksek Lisans Tezi; 2007.
[12] Pandey R. Introduction to Logistic Regression, https://www.quora.com/What-exactly-is-alogistic-regression-algorithm-in-machine-learning-What-are-its-applications (26.09.2019)
[13] Bozarık E. Sinir ağları ve derin öğrenme : Lojistik Regresyon, https://medium.com/deeplearning-turkiye/sinir-ağları-ve-derin-öğrenme-iii-lojistik-regresyon-cc9686981c6b(20.08.2019)