Predicting Workforce Demand in Service Organizations Using AI Techniques

  • Sara A. Sharif

Student thesis: Master's Thesis

Abstract

Prediction is about making claims of the future based on past information and current state [9]. Predicting resource demand for the future can help many service organizations to adjust their available resources, thus reach their goals of saving cost and enhance efficiency. In order to plan the organization's resources properly, it is important to be able to forecast the resources needed and their availability. It is vital for most service organizations to predict their future resource requirements to ensure that their upcoming projects are adequately staffed. In general resource prediction or forecasting has many key benefits, such as to ensure that the resources are being fully utilized and to help organization know the strengths and weaknesses of each project. Also, resource forecasting can help to meet demand for industry trends and reduce attrition rates [31]. This thesis project aims to design and develop applied prediction models that are capable of predicting resource demand for workforce management in several real world scenarios. A complete framework for predicting workforce demand in service organizations using different prediction techniques is provided. Several statistical methods for analyzing data such as filtering techniques, correlation analysis and clustering techniques are explained. Moreover, two case scenarios of two service organizations requiring forecasting of demand are discussed. They are 1) retail service with visual surveillance data and 2) telecom service with mobile workforce data. In both cases, an advanced neural network technique is used to produce forecast. The experiments were performed with the real-world data, and the results were compared against other popular techniques, such as linear regression and moving average which served as the techniques applied in past in respective organizations. The evaluation results of the proposed models are found to be promising where the accuracy of prediction is improved with the use of neural network. The key contribution of this project is to build prediction models for two real world use cases and report best accuracy achieved.
Date of AwardDec 2018
Original languageAmerican English

Keywords

  • Prediction
  • Linear Regression
  • Moving Average
  • Neural Network

Cite this

'