Implementación del modelo de regresión logística en el área comercial de W.C. Instalaciones

Abstract

Logistic regression represents a classification method used to anticipate the outcome of a variable categorized based on independent variables. By analyzing a data set composed of input variables and a binary outcome, it models the probability of occurrence of an event, commonly denoted as “yes” or “no.” Unlike linear regression, this approach accommodates the prediction of categorical outcomes and avoids assumptions about the normality of residuals. This statistical model was excutioned during an internship at Instalaciones Hidráulicas y Sanitarias W.C. S.A.S. where the development of tailored offers was developed at the request of potential clients in the commercial area. The company defines itself as an engineering entity specialized in providing services for the design and construction of hydraulic, sanitary, gas installations, etc. intended for residential, commercial, institutional and industrial buildings nationwide. Quarterly meetings were held in order to evaluate the performance of all areas through management indicators. The objective consisted on the data collection, cleansing and analysis in Python with the purpose of investigating its possible relationship with the current classification to offers. Data cleaning methods were used in the physical records, as well as in the Excel documents along with the creation of virtual tools for the registration for future entries. The process generated the presentation of a final report comprising the results obtained during the process.