Time series analysis and forecasting by example / Søren Bisgaard, Murat Kulahci

Includes bibliographical references and index

Table Of Contents: Preface 1 Time Series Data: Examples And Basic Concepts 1.1 Introduction 1.2 Examples of Time Series Data 1.3 Understanding Autocorrelation 1.4 The Wold Decomposition 1.5 The Impulse Response Function 1.6 Superposition Principle 1.7 Parsimonious Models Exercises 2 Visualizing Time Series Data Structures: Graphical Tools 2.1 Introduction 2.2 Graphical Analysis of Time Series 2.3 Graph Terminology 2.4 Graphical Perception 2.5 Principles of Graph Construction 2.6 Aspect Ratio 2.7 Time Series Plots 2.8 Bad Graphics Exercises 3 Stationary Models 3.1 Basics of Stationary Time Series Models 3.2 Autoregressive Moving Average (ARMA) Models 3.3 Stationarity and Invertibility of ARMA Models 3.4 Checking for Stationarity using Variogram 3.5 Transformation of Data Exercises 4 Nonstationary Models 4.1 Introduction 4.2 Detecting Nonstationarity 4.3 Autoregressive Integrated Moving Average (ARIMA) Models 4.4 Forecasting using ARIMA Models 4.5 Example 2: Concentration Measurements from a Chemical Process 4.6 The EWMA Forecast Exercises 5 Seasonal Models 5.1 Seasonal Data 5.2 Seasonal ARIMA Models 5.3 Forecasting using Seasonal ARIMA Models 5.4 Example 2: Company X's Sales Data Exercises 6 Time Series Model Selection 6.1 Introduction 6.2 Finding the "BEST" Model 6.3 Example: Internet Users Data 6.4 Model Selection Criteria 6.5 Impulse Response Function to Study the Differences in Models 6.6 Comparing Impulse Response Functions for Competing Models 6.7 ARIMA Models as Rational Approximations 6.8 AR Versus Arma Controversy 6.9 Final Thoughts on Model Selection Appendix 6.1 How to Compute Impulse Response Functions with a Spreadsheet Exercises 7 Additional Issues In Arima Models 7.1 Introduction 7.2 Linear Difference Equations 7.3 Eventual Forecast Function 7.4 Deterministic Trend Models 7.5 Yet Another Argument for Differencing 7.6 Constant Term in ARIMA Models 7.7 Cancellation of Terms in ARIMA Models 7.8 Stochastic Trend: Unit Root Nonstationary Processes 7.9 Overdifferencing and Underdifferencing 7.10 Missing Values in Time Series Data Exercises 8 Transfer Function Models 8.1 Introduction 8.2 Studying Input-Output Relationships 8.3 Example 1: The Box-Jenkins' Gas Furnace 8.4 Spurious Cross Correlations 8.5 Prewhitening 8.6 Identification of the Transfer Function 8.7 Modeling the Noise 8.8 The General Methodology for Transfer Function Models 8.9 Forecasting Using Transfer Function-Noise Models 8.10 Intervention Analysis Exercises 9 Additional Topics 9.1 Spurious Relationships 9.2 Autocorrelation in Regression 9.3 Process Regime Changes 9.4 Analysis of Multiple Time Series 9.5 Structural Analysis of Multiple Time Series Exercises Appendix A DATASETS USED IN THE EXAMPLES Table A.1 Temperature Readings from a Ceramic Furnace Table A.2 Chemical Process Temperature Readings Table A.3 Chemical Process Concentration Readings Table A.4 International Airline Passengers Table A.5 Company X's Sales Data Table A.6 Internet Users Data Table A.7 Historical Sea Level (mm) Data in Copenhagen, Denmark Table A.8 Gas Furnace Data Table A.9 Sales with Leading Indicator Table A.10 Crest/Colgate Market Share Table A.11 Simulated Process Data Table A.12 Coen and others (1969) Data Table A.13 Temperature Data from a Ceramic Furnace Table A.14 Temperature Readings from an Industrial Process Table A.15 US Hog Series Appendix B DATASETS USED IN THE EXERCISES Table B.1 Beverage Amount (ml) Table B.2 Pressure of the Steam Fed to a Distillation Column (bar) Table B.3 Number of Paper Checks Processed in a Local Bank Table B.4 Monthly Sea Levels in Los Angeles, California (mm) Table B.5 Temperature Readings from a Chemical Process (°C) Table B.6 Daily Average Exchange Rates between US Dollar and Euro Table B.7 Monthly US Unemployment Rates Table B.8 Monthly Residential Electricity Sales (MWh) and Average Residential Electricity Retail Price (c/kWh) in the United States Table B.9 Monthly Outstanding Consumer Credits Provided by Commercial Banks in the United States (million USD) Table B.10 100 Observations Simulated from an ARMA (1, 1) Process Table B.11 Quarterly Rental Vacancy Rates in the United States Table B.12 Wolfer Sunspot Numbers Table B.13 Viscosity Readings from a Chemical Process Table B.14 UK Midyear Population Table B.15 Unemployment and GDP data for the United Kingdom Table B.16 Monthly Crude Oil Production of OPEC Nations Table B.17 Quarterly Dollar Sales of Marshall Field and Company ([dollars]1000) Bibliography Index

Technology management scholar (U. of Massachusetts-Amherst) Bisgaard (1938-2010) and Kulahci (statistics, Technical U. of Denmark) found that many students and practitioners in statistics get frustrated trying to learn time series analysis, and either give up on it entirely or just plug data into a software package and accept what comes out. They set out to provide an introduction that is easy to understand and use, and that draws heavily from examples to demonstrate the principles and techniques. The profession of statistics needs at least a few people who know what is actually going on, they say, and who know the shortfalls of the statistical techniques being used. Annotation ©2011 Book News, Incorporated, Portland, OR (booknews.com)

An intuition-based approach enables you to master time series analysis with ease

Time Series Analysis and Forecasting by Example provides the fundamental techniques in time series analysis using various examples. By introducing necessary theory through examples that showcase the discussed topics, the authors successfully help readers develop an intuitive understanding of seemingly complicated time series models and their implications.

The book presents methodologies for time series analysis in a simplified, example-based approach. Using graphics, the authors discuss each presented example in detail and explain the relevant theory while also focusing on the interpretation of results in data analysis. Following a discussion of why autocorrelation is often observed when data is collected in time, subsequent chapters explore related topics, including:

Graphical tools in time series analysis Procedures for developing stationary, non-stationary, and seasonal models How to choose the best time series model Constant term and cancellation of terms in ARIMA models Forecasting using transfer function-noise models

The final chapter is dedicated to key topics such as spurious relationships, autocorrelation in regression, and multiple time series. Throughout the book, real-world examples illustrate step-by-step procedures and instructions using statistical software packages such as SAS®, JMP, Minitab, SCA, and R. A related Web site features PowerPoint slides to accompany each chapter as well as the book's data sets.

With its extensive use of graphics and examples to explain key concepts, Time Series Analysis and Forecasting by Example is an excellent book for courses on time series analysis at the upper-undergraduate and graduate levels. it also serves as a valuable resource for practitioners and researchers who carry out data and time series analysis in the fields of engineering, business, and economics.