Demand forecasting using Artificial Neural Network and demand pattern recognition at Sony Mobile Communication

Abstract

Title: Demand forecasting using Artificial Neural Network and demand pattern recognition at Sony Mobile Communication
Author: Christopher Duchateau

Background: Machine Learning is considered as one of the most important technologies in this era for the society and a possible fundamental driver of economic growth. Furthermore, machine learning technology is characterized to shape a new period in the economy. On this account, demand forecasting is a potential area for a successful application of Artificial Intelligence in order to benefit from the machine learning abilities. Understanding complex non-linear relationships in data, avoiding human made biases and improving the accuracy of the future demand are some aspects which are provided by machine learning technology. The impact of the new technologies could support traditional and coherent Supply Chain and Inventory Management planning decisions in a positive way.

Problem description: Demand forecasting is a time-consuming process and also combined with a high cost factor. Overestimation of demand results in excess stock keeping, which ties up capital. On other hand, underestimation of demand leads to unfulfilled orders, lost sales, reduced service levels and unhappy customers. Hence, more accurate demand forecasting enables a better planning process along the supply chain and help to avoid unnecessary costs for organizations.

Purpose: The purpose of the report is to investigate a forecasting model for linear and nonlinear time series using Artificial Neural Network algorithms and demand pattern recognition to improve the forecast accuracy.
Methodology: A single case study at Sony Mobile Communication was performed by following a data driven empirical approach. The purpose of the empiric research includes in particular the explanation of collected data within a developed model. Furthermore, the empiric approach recommends the application of quantitative and qualitative methods.

Conclusion: The overall conclusion of this project is that the developed model approach provides a higher forecast accuracy if traditional methods are leading to difficulties of identifying the exact demand trend pattern. The Artificial Neural Network is a robust and suitable method to classify demand trend pattern based on historical demand data. On the other hand, the chosen forecasting methods - ARIMA model for linear demand trends and the bootstrapping method for intermittent demand trends - allow a robust forecast of the future demand. The ability of the developed model is to differentiate between linear and non-linear models and to apply the right traditional forecasting method to reduce the error of miscalculation of the future demand. However, the selected forecasting methods are not improving the forecast accuracy compared to the traditional application of the ARIMA and bootstrapping method. The main aspect of the developed model can be explained by the suitability of ANNs to classify demand trends, whereas the prediction of the future demand can be improved further with more data and other algorithms. In order to use the ability of the machine learning technology, further research in the understanding of the customer behavior in combination weather, economic and marketing data needs to be done. In consideration of these aspects the machine learning technology offers interesting possibilities to improve demand forecasting.

Popular Abstract

Balancing demand and supply are one of the major challenges for various organizations. Overestimation of demand causes in extra stock keeping, which ties up excess inventory. On other hand, underestimation of demand leads to unfulfilled orders, lost sales and reduces the service level. Nowadays, customer behavior is influenced by many factors including, for example, marketing campaigns and weather, which makes it difficult for companies to predict the future demand. Demand forecasting is a field of predictive analytics, that aims to predict the future demand of customers. This is done by analyzing statistical data and identifying patterns and correlations.

Machine learning is considered as one of the most important new technologies for Big Data Analytics and a possible driver of economic growth. The ability of the technology offers possibilities to identify complex relationships in large data sets. Especially, the ability of the machine learning technology to handle large data volumes and to identify repeating patterns could support decision-making processes in organizations. As companies are relying more and more on the data they collect, traditional based forecasting methods are being replaced by data-powered and automated systems. Using big data demand prediction is enabling a wide range of organizations to leverage the potential of machine learning models.

In this study, a demand forecasting model has been developed to demonstrate an approach of using Artificial Neural Networks in the forecasting environment, where the aim was to reduce forecast errors and biases. Furthermore, the study investigated the necessary data requirements for a machine learning model. In combination with the machine learning technology – data quality and data volume are two of the major issues in the developing process of a forecasting model.

The results show that the Artificial Neural Network demonstrate a high accuracy of classifying different demand trends for a comparatively small data set and hence, reduce human made errors and biases. The model represents an initial step which needs further development in order to be suitable for a successful industrial application.

• author: Duchateau, Christopher
• supervisor: Johan Marklund
• organization: Production Management
• course: MIOM01 20191
• year: 2019
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• keywords: Machine Learning, Artificial Neural Intelligence, Demand Forecasting, Inventory Management, Control Chart Pattern Recognition
• language: English
• id: 8992753
• date added to LUP: 2019-10-15 14:35:49
• date last changed: 2019-10-15 14:35:49