Advanced Forecasting Techniques with Application to NN5 Time Series

Robert Andrawis

ISBN 10: 3659129623 / ISBN 13: 9783659129629

Published by LAP LAMBERT Academic Publishing Aug 2012, 2012

About this Item

Neuware -In this book we analyze the forecasting model that achieved the first rank in the Forecasting Competition for Artificial Neural Networks & Computational Intelligence NN5. The model is based on combination of machine learning and linear models. In addition, the approach and the experiments done to develop this model are explained in details to allow the reader to learn the methodology of developing such optimal models. The book also introduces a Bayesian forecasting approach for Holt's additive exponential smoothing method. Starting from the state space formulation, a formula for the forecast is derived and reduced to a two-dimensional integration that can be computed numerically in a straightforward way. In contrast with much of the work for exponential smoothing, this method produces the forecast density as well. The combinations of forecast are investigated as well in this book. A comparison between different combination methods is introduced with complete case study on tourism demand forecasting in Egypt.Books on Demand GmbH, Überseering 33, 22297 Hamburg 132 pp. Englisch.

Seller Inventory # 9783659129629

Bibliographic Details

• Title: Advanced Forecasting Techniques with Application to NN5 Time Series
• Publisher: LAP LAMBERT Academic Publishing Aug 2012
• Publication Date: 2012
• Language: English
• ISBN 10: 3659129623
• ISBN 13: 9783659129629
• Binding: Taschenbuch
• Condition: Neu
• Dimensions: 220W x 150H x 8D millimeters 220 millimeters width by 150 millimeters height by 8 millimeters depth
• Weight: 215 grams

About this title

Synopsis

In this book we analyze the forecasting model that achieved the first rank in the Forecasting Competition for Artificial Neural Networks & Computational Intelligence NN5. The model is based on combination of machine learning and linear models. In addition, the approach and the experiments done to develop this model are explained in details to allow the reader to learn the methodology of developing such optimal models. The book also introduces a Bayesian forecasting approach for Holt's additive exponential smoothing method. Starting from the state space formulation, a formula for the forecast is derived and reduced to a two-dimensional integration that can be computed numerically in a straightforward way. In contrast with much of the work for exponential smoothing, this method produces the forecast density as well. The combinations of forecast are investigated as well in this book. A comparison between different combination methods is introduced with complete case study on tourism demand forecasting in Egypt.

About the Author

Robert R. Andrawis, Received his B.S. and Master degree in 2006 and 2010 respectively from the Department of Computer Engineering, Cairo University, Egypt. He also worked in Egyptian Ministry of Communication, Valeo, Vodafone Egypt and Intel Egypt. He participated and obtained the first rank in the NN5 Forecasting Competition for Neural Networks.