¡Descarga forma normal ejerccios y más Resúmenes en PDF de Sistemática solo en Docsity!
UNIVERSIDAD ESAN
FACULTAD DE INGENIERÍA
INGENIERÍA INDUSTRIAL Y COMERCIAL
Propuesta de modelo predictivo empleando la técnica de machine learning para determinar la viabilidad de las cotizaciones de los proyectos de evacuación y señalización en la empresa P & R Arquitectos Consultores S.A.C. Trabajo de Suficiencia Profesional presentado en satisfacción parcial de los requerimientos para obtener el título profesional de Ingeniero (a) Industrial y Comercial AUTORES Ajalcriña Grimaldo, Lourdes Alexandra Alderete Arias, Fiorella Angelica Carrizales Valencia, Camila Graciela Tipe Carrasco, Jhosep Fernando ASESOR Fabian Arteaga, Junior ORCID Nº 0000-0001-9804- Marzo, 2023
ABSTRACT
We are in the era of digital transformation in which companies seek to automate their processes, this to improve productivity, customer experience, reduce costs and improve decision making. In this paper, we will analyze the company P & R Architects, which presents a problem in the evacuation and signaling projects since 76% of quotes are rejected. That is why a series of variables that are evaluated to determine the final status of the quotations were analyzed and, from them, Machine Learning models were built using 4 different techniques, such as: K-NN, Support Vector Machine, Logistic Regression and Decision Tree in order to obtain the most accurate model and, in addition, the results with normalization and without normalization were compared. From the results obtained, the decision tree technique has a better prediction of the quotes (85.88% with normalization). It is expected that the quote prediction model will help in making better decisions at the time of providing a final response to the client, either by adjusting in terms of costs or by negotiating directly with the client. Keywords: Machine Learning, quotations, logistic regression, K-NN, decision tree
