Please use this identifier to cite or link to this item:
Title: Applying statistical learning methods for forecasting prices and enhancing the probability of success in logistics tenders
Author: Tumkur Nataraj, Sachin
Álvarez Ruiz, Carlos
Sada Trabado, Lluna
Juan Pérez, Ángel Alejandro
Panadero Martínez, Javier
Bayliss, Christopher
Others: Universitat Oberta de Catalunya. Internet Interdisciplinary Institute (IN3)
Keywords: logistics
logistics tenders
reverse auctions
statistical learning methods
time series analysis
Issue Date: 25-Apr-2020
Publisher: Transportation Research Procedia
Citation: Nataraj, S., Alvarez, C., Sada, L., Juan, A.A, Panadero, J. & Bayliss, C. (2020). Applying statistical learning methods for forecasting prices and enhancing the probability of success in logistics tenders. Transportation Research Procedia, 47(), 529-536. doi: 10.1016/j.trpro.2020.03.128
Published in: Euro Working Group on Transportation Meeting (EWGT), Barcelona, 18-20 de setembre de 2019
Project identifier: info:eu-repo/grantAgreement/2018-1-ES01-KA103-049767
Also see:
Abstract: The concept of logistics tender or request for quotation is gaining importance among the logistics and transportation firms. Many long-term and long-distance transportation services are offered now under this type of reverse auction, and firms in the sector have to provide competitive prices if they want to win tenders. This paper explores the application of forecasting and statistical learning methods to enhance the competitiveness level of a firm when applying for tenders. On the one hand, time series analysis is used to: (i) forecast the long-term cost of the logistics service; and (ii) construct a 'risk-aware' interval for the prices to be offered in the bid. On the other hand, historical data is used to develop statistical learning models. These models are able to predict the probability of success in a given tender based on the actual values of different variables, including the service prices established in the previous stage. Some preliminary results are given, as well as a discussion on how these methods can be integrated into optimization models.
Language: English
ISSN: 2352-1465MIAR
Appears in Collections:Conference lectures

Files in This Item:
File Description SizeFormat 
  Restricted Access
709,15 kBAdobe PDFView/Open Request a copy