On time series data and optimal parameters

Size: px
Start display at page:

Download "On time series data and optimal parameters"

Transcription

1 Available online at Omega 32 (2004) On time series data and optimal parameters Rasmus Rasmussen Institute of Economics, Molde University College, P.O. Box 2110, Molde 6402, Norway Received 18 June 2001;accepted 26 September 2003 Abstract Forecasting using time series (TS) models are often based on linear regression or methods using various smoothing techniques. When estimating the parameters used in smoothing techniques, it has become a common practice to optimize the smoothing constants (parameters). This new practice is a result of the ease such methods can be accomplished when using the built in Solver optimization tool in modern spreadsheets. However, the capabilities of Solver can be utilized further to optimize more of the parameters, particularly the initial or starting parameters. This paper presents examples of exponential smoothing techniques, demonstrating improved ts when adopting this idea of optimizing the initial parameters as well as the smoothing constants. It also demonstrates that linear regression is a special case of Holt s exponential smoothing model with trend. Normalization of the seasonal parameters in models incorporating seasonality is also discussed, showing improved ts to TS data. Educators are encouraged to adopt the idea of letting Solver optimize more of the parameters than what is common practice today, in other models and in other elds.? 2003 Elsevier Ltd. All rights reserved. Keywords: Forecasting;Time series;regression;spreadsheets;education 1. Introduction Forecasting using time series (TS) analysis is a frequently used quantitative technique when numbers concerning the future are required. Spreadsheets used on an electronic computer are becoming a common tool for managers and analysts. Even though software specially designed for forecasting is available, the exibility and ease of use of spreadsheets combined with good graphing capabilities often makes them the choice of tools when performing the TS analysis [1]. Using spreadsheets to perform linear regression analysis or other models such as exponential smoothing techniques in TS analysis is becoming a standard in management science text books [2 5]. It is also becoming common practice to use Solver to choose the smoothing constant(s) [3, 5] for exponential smoothing techniques. Corresponding author. Fax: address: rasmus.rasmussen@himolde.no (R. Rasmussen). There are also add-ins to spreadsheets performing TS analysis, as CB Predictor, a tool that is a part of Crystal Ball Pro, 1 and ForecastX. 2 How the smoothing constants and initial values are computed in these add-ins are not always disclosed in their manuals. A numerical example seems to indicate that only the smoothing constants are optimized. In this paper I argue to go one step further, and also optimize the initial or starting parameters. These values are usually estimated based on the rst part of data in the TS. Often very simple formulas are used, as a shortcut to maintain the spreadsheet model simple. If the model includes seasonal adjustments, then normalization of the seasonal parameters often are skipped for the same reason. The normalization of the seasonal adjustments is seldom mentioned in these introductory text books. For short-range forecasts skipping normalization will normally not make the forecasts go out of proportions, and long-range forecasts is usually not advisable based on these models. But skipping 1 By Decisioneering; 2 By John Galt Solutions, Inc.; /$ - see front matter? 2003 Elsevier Ltd. All rights reserved. doi: /j.omega

2 112 R. Rasmussen / Omega 32 (2004) Fig. 1. Linear regression on a time series. normalization may cause systematic forecasting errors compared to the theoretical model [6]. This paper therefore suggests utilizing Solver to simultaneously optimize the smoothing constants and the initial values in the spreadsheet models forecasting TS data, when exponential smoothing methods are being used. With the increasing capabilities of Solver, 3 educators are encouraged to adopt the idea of optimizing parameters normally treated as constants, also in other types of models. As such, this paper is heavily inuenced by Ragsdale and Plane [1], advising to simultaneously optimize the regression parameters and seasonal adjustments when regression models are used to forecast TS. 2. Models with trend When a TS analysis with only trend is performed, a number of models can be applied. This paper focuses on exponential smoothing models, and illustrates Holt s model compared to linear regression. A linear regression model is easily implemented in a spreadsheet by using the built in TREND function to forecast a TS variable Y t (given historical data for the time 3 Premium Solver Platform v3.5 is used in this paper. Frontline Systems; periods t =1; 2; 3;:::;n). Forecasts based on linear regression are computed according to F t = b 0 + b 1t; (1) where the regression parameters b 0 and b 1 in (1) is estimated by minimizing the mean square error (MSE): Min : MSE (2) b 0 ;b 1 and the MSE in (2) is dened as MSE = 1 n (Y t F t) 2 : (3) n t=1 To reveal the parameters b 0 and b 1 the regression tool or Solver can be used, but the TREND function in Excel is suf- cient to make the forecasts directly. (Technically regression is minimizing the sum of squared error (SSE), which equals n MSE.) MSE for the complete series of data (t =1; 2; 3;:::;20) for the linear regression model in Fig. 1 is 3937, which is the value minimized by the TREND function. Some models use rst year for initialization, so forecasts can only be done for t = 5 20, and MSE for that period is therefore displayed for comparisons. MSE for the last cycle is also displayed, as this period often is regarded being most representative for the upcoming future. Holt s method is more exible than linear regression since it updates the level and trend parameters, which are constant

3 R. Rasmussen / Omega 32 (2004) Fig. 2. Holt s method with optimized smoothing parameters. in linear regression. Forecasts at time t for period t + k are made by F t+k = E t + kt t: (4) The level E t is updated as E t = Y t +(1 )[E t 1 + T t 1]: (5) The trend T t is updated as: T t = (E t E t 1)+(1 )T t 1: (6) Initial values have to be calculated to be able to update the values for level and trend. Several methods can be applied, but to keep the formulas in the spreadsheet simple, the following formulas are sometimes used: E 1 = Y 1; (7) T 1 =0: (8) The smoothing constants and are parameters set within the limits 0 6, 6 1. Solver can be used to minimize MSE using one period forecasts (k = 1) for periods 2 20, but for comparisons the periods 5 20 is minimized. The linear regression model in Fig. 1 has a better t to the data (lower MSE) than the implementation of Holt s method as shown in Fig. 2. This is caused by how the initial values for level and trend are computed. Even though more sophisticated forms for Eqs. (7) and (8) can be applied, the initial values for level and trend should be part of the optimization model to get the best possible t. Implementing this modication of the spreadsheet model, as in Fig. 3, shows that if the optimal values for the smoothing constants and = 0, then E 0 = b 0 and T 0 = b 1. In this case Holt s model equals simple linear regression. This implementation of Holt s model will never produce inferior ts compared to linear regression. (The Solver model will have the same objective, but more decision variables, of which some applies to similar variables.) 3. Models with trend and seasons Holt Winter s method adjusts for seasonal eects, and can be applied with both additive and multiplicative seasonal

4 114 R. Rasmussen / Omega 32 (2004) Fig. 3. Holt s method with optimized smoothing parameters and initial values. eects. First the multiplicative version will be illustrated. In this model, forecasts are adjusted for seasonal eects according to (9) F t+k =(E t + kt t)s t+k p : (9) In Holt Winter s method P equals the number of seasons per cycle (as 4 quarters per year). The updating procedure for the level is slightly modied [ ] Yt E t = +(1 )[E t 1 + T t 1]: (10) S t p The trend is updated as before and repeated T t = (E t E t 1)+(1 )T t 1: (11) The seasonal parameters are updated according to [ ] Yt S t = +(1 )S t p: (12) E t To start updating the level, trend and seasonal parameters, we have to estimate initial values. Again there are several options. To preserve as much as possible of the data series for testing forecasts, a simple procedure of using only the rst cycle as initial period will be used. Y t S t = (1=p p t=1 Yt) for t =1;:::;p; (13) E p = Yp ; S p (14) T p =0: (15) The smoothing constants ; and are parameters set within the limits 0 6,, 6 1. Solver can be used to minimize MSE for the periods This is implemented in Fig. 4. The TS data have obviously seasonal patterns, and the model incorporating seasonal eects makes a signicant improved t to the data compared to previous models with only trend. However, we can improve the t even further, by adopting the idea of optimizing the initial values for level, trend and seasonal parameters;in addition to optimizing the smoothing constants. In Fig. 5 a new initial cycle is added, and Solver is set to optimize the starting values for level, trend and seasons as well as the smoothing constants, and. For TS with limited amounts of historical data, this technique can be valuable. We are now able to use all observations for testing purposes, not having to devote some part of the TS data for initialization.

5 R. Rasmussen / Omega 32 (2004) Fig. 4. Holt Winter s multiplicative;with optimized smoothing parameters. The MSE is signicantly reduced when both the initial parameters and smoothing constants are optimized, in this example MSE is improved for all compared periods. If there are reasons to believe that recent data is more representative for the expected future than older observations, Solver can be set to minimize MSE for the last cycle(s) instead of for the whole data set. Or the oldest and no longer representative parts of the data set can be deleted. Since we do not have to reserve data for initialization, the latter may be the most ecient strategy. 4. Normalization A note on normalization of the seasonal parameters should be made. As can be seen in Fig. 5 the sum of the initial seasons s 3 +s 2 +s 1 +s 0 =9; 631. This should theoretically be equal P (4), or average to 1 for multiplicative models. If not, the model will make systematic forecasting errors [6]. Long-range forecasts can easily grow out of proportions. This is illustrated in the examples by the fact that this model is the one with the largest value for the latest forecast. Normalization is usually skipped in introductory textbooks, maybe for convenience, since the spreadsheet becomes quite more complex. Normalization involves an iterative procedure in updating the seasonal parameters. The average A 0 must be computed before updating A t = 1 P t i=t p+1 S i: (16) The normalized seasonal parameters for periods 0; 1;:::; p 1 are then computed at t =0as S i Si A t for i = t; t 1;:::;t p +1: (17) Then formulas (10) (12), (16) and (17) are applied for updating at t =1; 2;:::;n. The normalization will assure that the seasonal parameters do not adopt changes in level or trend, only adjust to changes in the seasonal variations. Again Solver is set to optimize the initial values for level, trend and seasons, along with the smoothing constants ; and. The result is presented in Fig. 6.

6 116 R. Rasmussen / Omega 32 (2004) Fig. 5. Holt Winter s multiplicative;with optimized smoothing parameters and initial values. The smoothing constants, and get the same optimal value as without normalization, and the forecasts are identical with the model in Fig. 5, where normalization is not performed. The only change from Fig. 5 is the values for the initial parameters. This occurs because the optimal smoothing constant for the seasonal parameters is equal zero in this example. In that case the rst normalization is sucient, since the updated values will not change. In the version with additive seasonal adjustments, the advantage of normalization is more profound. The forecasts are changed as follows F t+k = E t + kt t + S t+k p : (18) The updating procedure for the level, trend and seasons are slightly modied E t = (Y t S t p)+(1 )[E t 1 + T t 1]; (19) T t = (E t E t 1)+(1 )T t 1; (20) S t = (Y t E t)+(1 )S t p: (21) Solver is used to minimize MSE for the periods 1 20 by adjusting the smoothing constants, and within the limits 0 6,, 6 1, as well as the initial parameter values for level, trend and seasons. This is implemented in Fig. 7. The additive seasonal adjustments seem to not t the data as well as the multiplicative adjustments. But to take full advantage of the model we again should normalize the seasonal adjustments. This implementation is done in Fig. 8. The normalization in a model with additive adjustments of seasonal eects is similar to the multiplicative model. In an additive model the sum of the seasonal parameters for a cycle (year) should equal 0. The normalization will assure that the seasonal parameters do not adopt changes in level or trend, but only adjust to changes in the seasonal variations. The average A 0 is computed before updating as in (16). The normalized seasonal parameters for periods 0; 1;:::; p 1 are then computed at t =0as S i S i A t for i = t; t 1;:::;t p +1: (22) Thereafter formulas (19), (20), (21), (16) and (22) are applied for updating at t =1; 2;:::;n.

7 R. Rasmussen / Omega 32 (2004) Fig. 6. Holt Winter s multiplicative;normalized optimized smoothing parameters and initial values. An easy way to make the initial seasonal parameters is to add a constraint to Soiver. Make a constraint that corresponds to A 0 = 0 for models with additive seasons or a constraint A 0 = 1 for models with multiplicative seasons. In Fig. 7 without normalization the average of the seasonal parameters is 144, indicating that some of the trend and level adjustments is adopted in the adjustments of the seasonal eects. In Fig. 8 the normalization limits the seasonal adjustments to only capture seasonal eects, forcing the level and trend parameters to fully adopt these eects. The normalization has improved the t as measured by MSE for the period being minimized (periods 1 20), and also for the last year isolated (periods 17 20). But MSE for periods 5 20 seems to be increased by normalization. However, using Solver to minimize MSE for periods 5 20, then the MSE and forecasts are equal for both the non-normalized and the normalized version of the model with additive seasonal adjustments. If MSE for the last year (periods 17 20) is being minimized, both the model with multiplicative adjustments of seasonal eects and the model with additive adjustments make perfect ts to the data (MSE =0), both with and without normalization. (The models in Figs. 5 8.) This astonishingly good result is not a surprise when considering that the optimization model has 9 decision variables to t the forecasts for the 4 last quarters. But Solver superbly solves the complexity of the optimization problem. This perfect t is not achieved without also optimizing the initial parameters (as the model in Fig. 4). The forecasting add-in tool CB Predictor selects Holt Winter s multiplicative method as the method being able to t this TS data best. It does not t the data as well as the optimized normalized version presented in Fig. 6. These examples indicate that normalization can help to improve the ability to t forecasts to TS data. 5. Empirical tests In these examples focus has been on tting a forecasting model as close as possible to time series data. The ultimate test is of course to see how the forecasts t the future, but these tests can only be done retrospectively. Therefore it is common practise to do blind tests by splitting the data in two series, the rst part is used to t the model to the data, and the recent part of the time series is used as

8 118 R. Rasmussen / Omega 32 (2004) Fig. 7. Holt Winter s additive;with optimized smoothing parameters and initial values. a hold out test, to see how accurate the model forecasts the unknown future. The model that most accurately describes the holdout series is then selected for making the actual forecasts, after being ret to the complete time series. These previous examples may be regarded as this last step. The rationale behind this practice is the belief that the model that most accurately prolonged its tted pattern in the initial series into the holdout series also will be most accurate when prolonging the pattern from the complete data series into the real future. Just selecting the model that has the best t to the complete time series may result in selecting a model that overts the data incorporating random uctuations that do not repeat themselves. And empirical tests seem to indicate that this is what is actually happening statistically sophisticated models do not outperform simple forecasting techniques, the opposite may be the case [7]. 4 This fact has not stopped courses in advanced forecasting methods from being thought at the universities or papers being published in journals. Instead emphasis is being made that forecasts should not rely on a single method, they should be made as a composite of several dierent forecasting methods, also regarding qualitative methods. To test if the technique described in this paper overts the time series, a comparison to ARIMA models was done [8]. 5 The last year time series was used as holdout period. Two dierent ARIMA models were developed, using ForecastX. One model was developed by letting the software automatically select an ARIMA model, the other model was completely specied to the software. The specied model had a slightly better t to the initial data. ForecastX was also given the opportunity to freely select a forecasting model from its vast arsenal of techniques this led to the selection of Holt Winter with multiplicative season. MSE for these models are listed in Table 2, following the models developed after the guidelines in this article. Table 1 lists the classication of the exponential smoothing models used. 4 Particularly Chapter Data from Example 9.10.

9 R. Rasmussen / Omega 32 (2004) Fig. 8. Holt Winter s additive;normalized and optimized smoothing parameters and initial values. Table 1 Pegel s classication Season Trend None Additive Multiplicative None A1 A2 A3 Additive B1 B2 B3 Multiplicative C1 C2 C3 Table 2 Using holdout data MSE Initial Holdout B2 Normalized 803,8 641,1 B3 Normalized 609,8 886,1 C2 Normalized 831,3 786,8 C3 Normalized 652,2 951,4 ARIMA(1,1,0) * (2,1,0) 771,9 1988,2 ARIMA(1,1,0) * (0,1,2) 719,0 1859,9 ForecastX B3 738,8 1287,5 The model that best ts the initial series is Holt Winter with additive trend and multiplicative season (B3). Based on the holdout series the selection is Holt Winter with additive trend and additive season (B2). We see that all models listed are making approximately equal good t to the data seen by the model (MSE range from ). The models made in accordance with this article have MSE in the holdout series comparable to the MSE in the initial series, two has lower and two has higher MSE in the holdout series. This is quite opposite to the more sophisticated statistical ARIMA models both have more that twice as large MSE in the holdout series than in the initial series. We see that the Holt Winter method applied by ForecastX (which does not normalize the seasons or apply the technique of simultaneously optimize the smoothing constants and the initial parameter values);also make a considerable larger MSE in the holdout series than in the initial series. This indicates that the recommendations made in this article do not lead to overt, comparing with the same model done the traditional way. The result is also in line with previous empirical tests more sophisticated (ARIMA) models do not outperform simpler (exponentially smoothing) models. Compared to the sophisticated models there is no

10 120 R. Rasmussen / Omega 32 (2004) indication that the methods described in this article will lead to greater overt, the opposite seem to be the case. 6. Optimization tools The optimization models for minimizing MSE will be non-linear due to a non-linear objective function. But adding the initial parameters to the decision variables also will create non-linearity, since decision variables will be multiplied with each other. With several decision variables in such complicated non-linear models, the optimization task will not be straight forward to solve. The standard Solver will nd the nearest local optimal solution relative to the initial values of the decision variables set by the spreadsheet user. To search for alternative local optimal solutions will be like looking for a needle in a haystack. Fortunately, the Premium Solver Platform has a multistart search option, which has proven quite ecient in these examples. 7. Conclusions This paper outlines an approach whereby smoothing and initial parameters are found by solving directly the problem of minimizing MSE. The advantage is not only better t to TS data. It will also reduce the need for data, since there will be no need to use data to estimate the initial parameters. When initial parameters are optimized along with the smoothing constants, linear regression becomes a special case of Holt s method. The paper also advice to normalize seasonal adjustments in forecasting models. This extra complexity pays o in better ts and a theoretical sounder model. References [1] Ragsdale CT, Plane DR. On modeling time series data using spreadsheets. Omega 2000;28(2): [2] Camm, JD, Evans JR, Management science and decision technology, 1st ed. Cincinnati, OH: Southwestern College Publishing;2000. p [3] Ragsdale, CT. Spreadsheet modeling and decision analysis, 3rd ed. Cincinnati, OH: Southwestern College Publishing;2001. p [4] Render B, Stair RM. Quantitative analysis for management, 7th ed. Englewood Clis, NJ: Prentice-Hall;2000. [5] Winston WL, Albright CS. Practical management science, 2nd ed. Pacic Grove, CA: Duxbury;2001. p [6] Mossin J. Operasjonsanalytiske emner. Oslo: Johan Grundt Tanum: s. [7] Makridakis, SG, Wheelwright SC, Hyndman RJ. Forecasting methods and applications, 3rd ed. New York: John Wiley & Sons Inc.;1998. p [8] Hanke JE, Wichern DW, Reitsch AG. Business forecasting, 7th ed. Upper Saddle River, NJ: Prentice Hall, Inc.;2001. p. 498.

INCREASING FORECASTING ACCURACY OF TREND DEMAND BY NON-LINEAR OPTIMIZATION OF THE SMOOTHING CONSTANT

INCREASING FORECASTING ACCURACY OF TREND DEMAND BY NON-LINEAR OPTIMIZATION OF THE SMOOTHING CONSTANT 58 INCREASING FORECASTING ACCURACY OF TREND DEMAND BY NON-LINEAR OPTIMIZATION OF THE SMOOTHING CONSTANT Sudipa Sarker 1 * and Mahbub Hossain 2 1 Department of Industrial and Production Engineering Bangladesh

More information

USING SEASONAL AND CYCLICAL COMPONENTS IN LEAST SQUARES FORECASTING MODELS

USING SEASONAL AND CYCLICAL COMPONENTS IN LEAST SQUARES FORECASTING MODELS Using Seasonal and Cyclical Components in Least Squares Forecasting models USING SEASONAL AND CYCLICAL COMPONENTS IN LEAST SQUARES FORECASTING MODELS Frank G. Landram, West Texas A & M University Amjad

More information

2) The three categories of forecasting models are time series, quantitative, and qualitative. 2)

2) The three categories of forecasting models are time series, quantitative, and qualitative. 2) Exam Name TRUE/FALSE. Write 'T' if the statement is true and 'F' if the statement is false. 1) Regression is always a superior forecasting method to exponential smoothing, so regression should be used

More information

Demand Forecasting When a product is produced for a market, the demand occurs in the future. The production planning cannot be accomplished unless

Demand Forecasting When a product is produced for a market, the demand occurs in the future. The production planning cannot be accomplished unless Demand Forecasting When a product is produced for a market, the demand occurs in the future. The production planning cannot be accomplished unless the volume of the demand known. The success of the business

More information

Time series Forecasting using Holt-Winters Exponential Smoothing

Time series Forecasting using Holt-Winters Exponential Smoothing Time series Forecasting using Holt-Winters Exponential Smoothing Prajakta S. Kalekar(04329008) Kanwal Rekhi School of Information Technology Under the guidance of Prof. Bernard December 6, 2004 Abstract

More information

How To Plan A Pressure Container Factory

How To Plan A Pressure Container Factory ScienceAsia 27 (2) : 27-278 Demand Forecasting and Production Planning for Highly Seasonal Demand Situations: Case Study of a Pressure Container Factory Pisal Yenradee a,*, Anulark Pinnoi b and Amnaj Charoenthavornying

More information

Moving averages. Rob J Hyndman. November 8, 2009

Moving averages. Rob J Hyndman. November 8, 2009 Moving averages Rob J Hyndman November 8, 009 A moving average is a time series constructed by taking averages of several sequential values of another time series. It is a type of mathematical convolution.

More information

Time Series and Forecasting

Time Series and Forecasting Chapter 22 Page 1 Time Series and Forecasting A time series is a sequence of observations of a random variable. Hence, it is a stochastic process. Examples include the monthly demand for a product, the

More information

Industry Environment and Concepts for Forecasting 1

Industry Environment and Concepts for Forecasting 1 Table of Contents Industry Environment and Concepts for Forecasting 1 Forecasting Methods Overview...2 Multilevel Forecasting...3 Demand Forecasting...4 Integrating Information...5 Simplifying the Forecast...6

More information

Time Series Forecasting Techniques

Time Series Forecasting Techniques 03-Mentzer (Sales).qxd 11/2/2004 11:33 AM Page 73 3 Time Series Forecasting Techniques Back in the 1970s, we were working with a company in the major home appliance industry. In an interview, the person

More information

Spreadsheets have become the principal software application for teaching decision models in most business

Spreadsheets have become the principal software application for teaching decision models in most business Vol. 8, No. 2, January 2008, pp. 89 95 issn 1532-0545 08 0802 0089 informs I N F O R M S Transactions on Education Teaching Note Some Practical Issues with Excel Solver: Lessons for Students and Instructors

More information

Time series forecasting

Time series forecasting Time series forecasting 1 The latest version of this document and related examples are found in http://myy.haaga-helia.fi/~taaak/q Time series forecasting The objective of time series methods is to discover

More information

Introduction to Logistic Regression

Introduction to Logistic Regression OpenStax-CNX module: m42090 1 Introduction to Logistic Regression Dan Calderon This work is produced by OpenStax-CNX and licensed under the Creative Commons Attribution License 3.0 Abstract Gives introduction

More information

MGT 267 PROJECT. Forecasting the United States Retail Sales of the Pharmacies and Drug Stores. Done by: Shunwei Wang & Mohammad Zainal

MGT 267 PROJECT. Forecasting the United States Retail Sales of the Pharmacies and Drug Stores. Done by: Shunwei Wang & Mohammad Zainal MGT 267 PROJECT Forecasting the United States Retail Sales of the Pharmacies and Drug Stores Done by: Shunwei Wang & Mohammad Zainal Dec. 2002 The retail sale (Million) ABSTRACT The present study aims

More information

BUS 655/655/455 Forecasting & Predictive Analytics How to Predict the Future with Confidence

BUS 655/655/455 Forecasting & Predictive Analytics How to Predict the Future with Confidence BUS 655/655/455 Forecasting & Predictive Analytics How to Predict the Future with Confidence Spring 2015 Wednesday 6:30-9:15, Room 208 Prof. Stephen P. Stuk, Ph.D. Office GBS 409 Mobile: 678-468-3452 use

More information

Modelling and Forecasting Packaged Food Product Sales Using Mathematical Programming

Modelling and Forecasting Packaged Food Product Sales Using Mathematical Programming Modelling and Forecasting Packaged Food Product Sales Using Mathematical Programming Saurabh Gupta 1, Nishant Kumar 2 saurabhgupta2dams@gmail.com Abstract Sales forecasting is one of the most common phenomena

More information

Forecasting Methods / Métodos de Previsão Week 1

Forecasting Methods / Métodos de Previsão Week 1 Forecasting Methods / Métodos de Previsão Week 1 ISCTE - IUL, Gestão, Econ, Fin, Contab. Diana Aldea Mendes diana.mendes@iscte.pt February 3, 2011 DMQ, ISCTE-IUL (diana.mendes@iscte.pt) Forecasting Methods

More information

2. What is the general linear model to be used to model linear trend? (Write out the model) = + + + or

2. What is the general linear model to be used to model linear trend? (Write out the model) = + + + or Simple and Multiple Regression Analysis Example: Explore the relationships among Month, Adv.$ and Sales $: 1. Prepare a scatter plot of these data. The scatter plots for Adv.$ versus Sales, and Month versus

More information

Fuzzy regression model with fuzzy input and output data for manpower forecasting

Fuzzy regression model with fuzzy input and output data for manpower forecasting Fuzzy Sets and Systems 9 (200) 205 23 www.elsevier.com/locate/fss Fuzzy regression model with fuzzy input and output data for manpower forecasting Hong Tau Lee, Sheu Hua Chen Department of Industrial Engineering

More information

Objectives of Chapters 7,8

Objectives of Chapters 7,8 Objectives of Chapters 7,8 Planning Demand and Supply in a SC: (Ch7, 8, 9) Ch7 Describes methodologies that can be used to forecast future demand based on historical data. Ch8 Describes the aggregate planning

More information

IDENTIFICATION OF DEMAND FORECASTING MODEL CONSIDERING KEY FACTORS IN THE CONTEXT OF HEALTHCARE PRODUCTS

IDENTIFICATION OF DEMAND FORECASTING MODEL CONSIDERING KEY FACTORS IN THE CONTEXT OF HEALTHCARE PRODUCTS IDENTIFICATION OF DEMAND FORECASTING MODEL CONSIDERING KEY FACTORS IN THE CONTEXT OF HEALTHCARE PRODUCTS Sushanta Sengupta 1, Ruma Datta 2 1 Tata Consultancy Services Limited, Kolkata 2 Netaji Subhash

More information

IBM SPSS Forecasting 22

IBM SPSS Forecasting 22 IBM SPSS Forecasting 22 Note Before using this information and the product it supports, read the information in Notices on page 33. Product Information This edition applies to version 22, release 0, modification

More information

TIME SERIES ANALYSIS & FORECASTING

TIME SERIES ANALYSIS & FORECASTING CHAPTER 19 TIME SERIES ANALYSIS & FORECASTING Basic Concepts 1. Time Series Analysis BASIC CONCEPTS AND FORMULA The term Time Series means a set of observations concurring any activity against different

More information

Forecasting in STATA: Tools and Tricks

Forecasting in STATA: Tools and Tricks Forecasting in STATA: Tools and Tricks Introduction This manual is intended to be a reference guide for time series forecasting in STATA. It will be updated periodically during the semester, and will be

More information

Chapter 27 Using Predictor Variables. Chapter Table of Contents

Chapter 27 Using Predictor Variables. Chapter Table of Contents Chapter 27 Using Predictor Variables Chapter Table of Contents LINEAR TREND...1329 TIME TREND CURVES...1330 REGRESSORS...1332 ADJUSTMENTS...1334 DYNAMIC REGRESSOR...1335 INTERVENTIONS...1339 TheInterventionSpecificationWindow...1339

More information

Association Between Variables

Association Between Variables Contents 11 Association Between Variables 767 11.1 Introduction............................ 767 11.1.1 Measure of Association................. 768 11.1.2 Chapter Summary.................... 769 11.2 Chi

More information

CB Predictor 1.6. User Manual

CB Predictor 1.6. User Manual CB Predictor 1.6 User Manual This manual, and the software described in it, are furnished under license and may only be used or copied in accordance with the terms of the license agreement. Information

More information

A Generalized PERT/CPM Implementation in a Spreadsheet

A Generalized PERT/CPM Implementation in a Spreadsheet A Generalized PERT/CPM Implementation in a Spreadsheet Abstract Kala C. Seal College of Business Administration Loyola Marymount University Los Angles, CA 90045, USA kseal@lmumail.lmu.edu This paper describes

More information

Evaluation of forecasting methods for intermittent parts demand in the eld of aviation: a predictive model

Evaluation of forecasting methods for intermittent parts demand in the eld of aviation: a predictive model Available online at www.sciencedirect.com Computers & Operations Research 30 (2003) 2097 2114 www.elsevier.com/locate/dsw Evaluation of forecasting methods for intermittent parts demand in the eld of aviation:

More information

Theory at a Glance (For IES, GATE, PSU)

Theory at a Glance (For IES, GATE, PSU) 1. Forecasting Theory at a Glance (For IES, GATE, PSU) Forecasting means estimation of type, quantity and quality of future works e.g. sales etc. It is a calculated economic analysis. 1. Basic elements

More information

Process Simulation in Excel for a Quantitative Management Course

Process Simulation in Excel for a Quantitative Management Course Abstract Process Simulation in Excel for a Quantitative Management Course Raymond R. Hill Department of Operational Sciences Air Force Institute of Technology Wright-Patterson AFB, OH 45424 ray.hill@afit.edu

More information

TIME SERIES ANALYSIS

TIME SERIES ANALYSIS TIME SERIES ANALYSIS Ramasubramanian V. I.A.S.R.I., Library Avenue, New Delhi- 110 012 ram_stat@yahoo.co.in 1. Introduction A Time Series (TS) is a sequence of observations ordered in time. Mostly these

More information

Forecasting Methods. What is forecasting? Why is forecasting important? How can we evaluate a future demand? How do we make mistakes?

Forecasting Methods. What is forecasting? Why is forecasting important? How can we evaluate a future demand? How do we make mistakes? Forecasting Methods What is forecasting? Why is forecasting important? How can we evaluate a future demand? How do we make mistakes? Prod - Forecasting Methods Contents. FRAMEWORK OF PLANNING DECISIONS....

More information

International College of Economics and Finance Syllabus Probability Theory and Introductory Statistics

International College of Economics and Finance Syllabus Probability Theory and Introductory Statistics International College of Economics and Finance Syllabus Probability Theory and Introductory Statistics Lecturer: Mikhail Zhitlukhin. 1. Course description Probability Theory and Introductory Statistics

More information

Two-step competition process leads to quasi power-law income distributions Application to scientic publication and citation distributions

Two-step competition process leads to quasi power-law income distributions Application to scientic publication and citation distributions Physica A 298 (21) 53 536 www.elsevier.com/locate/physa Two-step competition process leads to quasi power-law income distributions Application to scientic publication and citation distributions Anthony

More information

Measurement with Ratios

Measurement with Ratios Grade 6 Mathematics, Quarter 2, Unit 2.1 Measurement with Ratios Overview Number of instructional days: 15 (1 day = 45 minutes) Content to be learned Use ratio reasoning to solve real-world and mathematical

More information

Numerical integration of a function known only through data points

Numerical integration of a function known only through data points Numerical integration of a function known only through data points Suppose you are working on a project to determine the total amount of some quantity based on measurements of a rate. For example, you

More information

Readers will be provided a link to download the software and Excel files that are used in the book after payment. Please visit http://www.xlpert.

Readers will be provided a link to download the software and Excel files that are used in the book after payment. Please visit http://www.xlpert. Readers will be provided a link to download the software and Excel files that are used in the book after payment. Please visit http://www.xlpert.com for more information on the book. The Excel files are

More information

Production Planning. Chapter 4 Forecasting. Overview. Overview. Chapter 04 Forecasting 1. 7 Steps to a Forecast. What is forecasting?

Production Planning. Chapter 4 Forecasting. Overview. Overview. Chapter 04 Forecasting 1. 7 Steps to a Forecast. What is forecasting? Chapter 4 Forecasting Production Planning MRP Purchasing Sales Forecast Aggregate Planning Master Production Schedule Production Scheduling Production What is forecasting? Types of forecasts 7 steps of

More information

CHAPTER 11 FORECASTING AND DEMAND PLANNING

CHAPTER 11 FORECASTING AND DEMAND PLANNING OM CHAPTER 11 FORECASTING AND DEMAND PLANNING DAVID A. COLLIER AND JAMES R. EVANS 1 Chapter 11 Learning Outcomes l e a r n i n g o u t c o m e s LO1 Describe the importance of forecasting to the value

More information

ISSUES IN UNIVARIATE FORECASTING

ISSUES IN UNIVARIATE FORECASTING ISSUES IN UNIVARIATE FORECASTING By Rohaiza Zakaria 1, T. Zalizam T. Muda 2 and Suzilah Ismail 3 UUM College of Arts and Sciences, Universiti Utara Malaysia 1 rhz@uum.edu.my, 2 zalizam@uum.edu.my and 3

More information

A Comparative Study of the Pickup Method and its Variations Using a Simulated Hotel Reservation Data

A Comparative Study of the Pickup Method and its Variations Using a Simulated Hotel Reservation Data A Comparative Study of the Pickup Method and its Variations Using a Simulated Hotel Reservation Data Athanasius Zakhary, Neamat El Gayar Faculty of Computers and Information Cairo University, Giza, Egypt

More information

Linear Programming. Solving LP Models Using MS Excel, 18

Linear Programming. Solving LP Models Using MS Excel, 18 SUPPLEMENT TO CHAPTER SIX Linear Programming SUPPLEMENT OUTLINE Introduction, 2 Linear Programming Models, 2 Model Formulation, 4 Graphical Linear Programming, 5 Outline of Graphical Procedure, 5 Plotting

More information

8. Time Series and Prediction

8. Time Series and Prediction 8. Time Series and Prediction Definition: A time series is given by a sequence of the values of a variable observed at sequential points in time. e.g. daily maximum temperature, end of day share prices,

More information

Outline. Role of Forecasting. Characteristics of Forecasts. Logistics and Supply Chain Management. Demand Forecasting

Outline. Role of Forecasting. Characteristics of Forecasts. Logistics and Supply Chain Management. Demand Forecasting Logistics and Supply Chain Management Demand Forecasting 1 Outline The role of forecasting in a supply chain Characteristics ti of forecasts Components of forecasts and forecasting methods Basic approach

More information

Call Centre Helper - Forecasting Excel Template

Call Centre Helper - Forecasting Excel Template Call Centre Helper - Forecasting Excel Template This is a monthly forecaster, and to use it you need to have at least 24 months of data available to you. Using the Forecaster Open the spreadsheet and enable

More information

one Introduction chapter OVERVIEW CHAPTER

one Introduction chapter OVERVIEW CHAPTER one Introduction CHAPTER chapter OVERVIEW 1.1 Introduction to Decision Support Systems 1.2 Defining a Decision Support System 1.3 Decision Support Systems Applications 1.4 Textbook Overview 1.5 Summary

More information

Dynamic stang in a telephone call center aiming to immediately answer all calls

Dynamic stang in a telephone call center aiming to immediately answer all calls Operations Research Letters 24 (1999) 25 212 www.elsevier.com/locate/orms Dynamic stang in a telephone call center aiming to immediately answer all calls Ward Whitt Room A117, AT&T Labs, Shannon Laboratory,

More information

Demand forecasting & Aggregate planning in a Supply chain. Session Speaker Prof.P.S.Satish

Demand forecasting & Aggregate planning in a Supply chain. Session Speaker Prof.P.S.Satish Demand forecasting & Aggregate planning in a Supply chain Session Speaker Prof.P.S.Satish 1 Introduction PEMP-EMM2506 Forecasting provides an estimate of future demand Factors that influence demand and

More information

Forecasting methods applied to engineering management

Forecasting methods applied to engineering management Forecasting methods applied to engineering management Áron Szász-Gábor Abstract. This paper presents arguments for the usefulness of a simple forecasting application package for sustaining operational

More information

RELEVANT TO ACCA QUALIFICATION PAPER P3. Studying Paper P3? Performance objectives 7, 8 and 9 are relevant to this exam

RELEVANT TO ACCA QUALIFICATION PAPER P3. Studying Paper P3? Performance objectives 7, 8 and 9 are relevant to this exam RELEVANT TO ACCA QUALIFICATION PAPER P3 Studying Paper P3? Performance objectives 7, 8 and 9 are relevant to this exam Business forecasting and strategic planning Quantitative data has always been supplied

More information

Time Series Analysis: Basic Forecasting.

Time Series Analysis: Basic Forecasting. Time Series Analysis: Basic Forecasting. As published in Benchmarks RSS Matters, April 2015 http://web3.unt.edu/benchmarks/issues/2015/04/rss-matters Jon Starkweather, PhD 1 Jon Starkweather, PhD jonathan.starkweather@unt.edu

More information

OBJECTIVE ASSESSMENT OF FORECASTING ASSIGNMENTS USING SOME FUNCTION OF PREDICTION ERRORS

OBJECTIVE ASSESSMENT OF FORECASTING ASSIGNMENTS USING SOME FUNCTION OF PREDICTION ERRORS OBJECTIVE ASSESSMENT OF FORECASTING ASSIGNMENTS USING SOME FUNCTION OF PREDICTION ERRORS CLARKE, Stephen R. Swinburne University of Technology Australia One way of examining forecasting methods via assignments

More information

Forecasting in supply chains

Forecasting in supply chains 1 Forecasting in supply chains Role of demand forecasting Effective transportation system or supply chain design is predicated on the availability of accurate inputs to the modeling process. One of the

More information

Outline: Demand Forecasting

Outline: Demand Forecasting Outline: Demand Forecasting Given the limited background from the surveys and that Chapter 7 in the book is complex, we will cover less material. The role of forecasting in the chain Characteristics of

More information

UNDERSTANDING ANALYSIS OF COVARIANCE (ANCOVA)

UNDERSTANDING ANALYSIS OF COVARIANCE (ANCOVA) UNDERSTANDING ANALYSIS OF COVARIANCE () In general, research is conducted for the purpose of explaining the effects of the independent variable on the dependent variable, and the purpose of research design

More information

STRUTS: Statistical Rules of Thumb. Seattle, WA

STRUTS: Statistical Rules of Thumb. Seattle, WA STRUTS: Statistical Rules of Thumb Gerald van Belle Departments of Environmental Health and Biostatistics University ofwashington Seattle, WA 98195-4691 Steven P. Millard Probability, Statistics and Information

More information

White Paper. Desktop Forecasting for Small and Midsize Businesses

White Paper. Desktop Forecasting for Small and Midsize Businesses White Paper Desktop Forecasting for Small and Midsize Businesses Contents Introduction... 1 Why Forecasting Is Important... 1 Looking Inside SAS Forecasting for Desktop... 1 Introduction... 1 Overview

More information

Journal Of Financial And Strategic Decisions Volume 10 Number 2 Summer 1997

Journal Of Financial And Strategic Decisions Volume 10 Number 2 Summer 1997 Journal Of Financial And Strategic Decisions Volume 10 Number 2 Summer 1997 AN EMPIRICAL INVESTIGATION OF PUT OPTION PRICING: A SPECIFICATION TEST OF AT-THE-MONEY OPTION IMPLIED VOLATILITY Hongshik Kim,

More information

Math 132. Population Growth: the World

Math 132. Population Growth: the World Math 132 Population Growth: the World S. R. Lubkin Application If you think growth in Raleigh is a problem, think a little bigger. The population of the world has been growing spectacularly fast in the

More information

CHAPTER 6 FINANCIAL FORECASTING

CHAPTER 6 FINANCIAL FORECASTING TUTORIAL NOTES CHAPTER 6 FINANCIAL FORECASTING 6.1 INTRODUCTION Forecasting represents an integral part of any planning process that is undertaken by all firms. Firms must make decisions today that will

More information

Data analysis in supersaturated designs

Data analysis in supersaturated designs Statistics & Probability Letters 59 (2002) 35 44 Data analysis in supersaturated designs Runze Li a;b;, Dennis K.J. Lin a;b a Department of Statistics, The Pennsylvania State University, University Park,

More information

Chapter Seven. Multiple regression An introduction to multiple regression Performing a multiple regression on SPSS

Chapter Seven. Multiple regression An introduction to multiple regression Performing a multiple regression on SPSS Chapter Seven Multiple regression An introduction to multiple regression Performing a multiple regression on SPSS Section : An introduction to multiple regression WHAT IS MULTIPLE REGRESSION? Multiple

More information

Using EXCEL Solver October, 2000

Using EXCEL Solver October, 2000 Using EXCEL Solver October, 2000 2 The Solver option in EXCEL may be used to solve linear and nonlinear optimization problems. Integer restrictions may be placed on the decision variables. Solver may be

More information

Creating, Solving, and Graphing Systems of Linear Equations and Linear Inequalities

Creating, Solving, and Graphing Systems of Linear Equations and Linear Inequalities Algebra 1, Quarter 2, Unit 2.1 Creating, Solving, and Graphing Systems of Linear Equations and Linear Inequalities Overview Number of instructional days: 15 (1 day = 45 60 minutes) Content to be learned

More information

Al-Jo anee Company: support department cost allocations with matrices to improve decision making

Al-Jo anee Company: support department cost allocations with matrices to improve decision making Al-Jo anee Company: support department cost allocations with matrices to improve decision making ABSTRACT Saad S. Hussein University of Tikrit, Iraq Dennis F. Togo University of New Mexico, USA The direct,

More information

16 : Demand Forecasting

16 : Demand Forecasting 16 : Demand Forecasting 1 Session Outline Demand Forecasting Subjective methods can be used only when past data is not available. When past data is available, it is advisable that firms should use statistical

More information

A New Method for Electric Consumption Forecasting in a Semiconductor Plant

A New Method for Electric Consumption Forecasting in a Semiconductor Plant A New Method for Electric Consumption Forecasting in a Semiconductor Plant Prayad Boonkham 1, Somsak Surapatpichai 2 Spansion Thailand Limited 229 Moo 4, Changwattana Road, Pakkred, Nonthaburi 11120 Nonthaburi,

More information

Applicability and accuracy of quantitative forecasting models applied in actual firms A case study at The Company

Applicability and accuracy of quantitative forecasting models applied in actual firms A case study at The Company Applicability and accuracy of quantitative forecasting models applied in actual firms A case study at The Company Master of Science Thesis in the Management and Economics of Innovation Programme JOHAN

More information

Teaching model: C1 a. General background: 50% b. Theory-into-practice/developmental 50% knowledge-building: c. Guided academic activities:

Teaching model: C1 a. General background: 50% b. Theory-into-practice/developmental 50% knowledge-building: c. Guided academic activities: 1. COURSE DESCRIPTION Degree: Double Degree: Derecho y Finanzas y Contabilidad (English teaching) Course: STATISTICAL AND ECONOMETRIC METHODS FOR FINANCE (Métodos Estadísticos y Econométricos en Finanzas

More information

ADVANCED FORECASTING MODELS USING SAS SOFTWARE

ADVANCED FORECASTING MODELS USING SAS SOFTWARE ADVANCED FORECASTING MODELS USING SAS SOFTWARE Girish Kumar Jha IARI, Pusa, New Delhi 110 012 gjha_eco@iari.res.in 1. Transfer Function Model Univariate ARIMA models are useful for analysis and forecasting

More information

Big Ideas in Mathematics

Big Ideas in Mathematics Big Ideas in Mathematics which are important to all mathematics learning. (Adapted from the NCTM Curriculum Focal Points, 2006) The Mathematics Big Ideas are organized using the PA Mathematics Standards

More information

TIME SERIES ANALYSIS

TIME SERIES ANALYSIS TIME SERIES ANALYSIS L.M. BHAR AND V.K.SHARMA Indian Agricultural Statistics Research Institute Library Avenue, New Delhi-0 02 lmb@iasri.res.in. Introduction Time series (TS) data refers to observations

More information

ICASL - Business School Programme

ICASL - Business School Programme ICASL - Business School Programme Quantitative Techniques for Business (Module 3) Financial Mathematics TUTORIAL 2A This chapter deals with problems related to investing money or capital in a business

More information

Mode and Patient-mix Adjustment of the CAHPS Hospital Survey (HCAHPS)

Mode and Patient-mix Adjustment of the CAHPS Hospital Survey (HCAHPS) Mode and Patient-mix Adjustment of the CAHPS Hospital Survey (HCAHPS) April 30, 2008 Abstract A randomized Mode Experiment of 27,229 discharges from 45 hospitals was used to develop adjustments for the

More information

Simple linear regression

Simple linear regression Simple linear regression Introduction Simple linear regression is a statistical method for obtaining a formula to predict values of one variable from another where there is a causal relationship between

More information

Forecasting method selection in a global supply chain

Forecasting method selection in a global supply chain Forecasting method selection in a global supply chain Everette S. Gardner, Jr. * Department of Decision and Information Sciences C.T. Bauer College of Business 334 Melcher Hall, Houston, Texas USA 77204-6021

More information

A Quantitative Approach to Commercial Damages. Applying Statistics to the Measurement of Lost Profits + Website

A Quantitative Approach to Commercial Damages. Applying Statistics to the Measurement of Lost Profits + Website Brochure More information from http://www.researchandmarkets.com/reports/2212877/ A Quantitative Approach to Commercial Damages. Applying Statistics to the Measurement of Lost Profits + Website Description:

More information

Forecasting irregular demand for spare parts inventory

Forecasting irregular demand for spare parts inventory Forecasting irregular demand for spare parts inventory Dang Quang Vinh Department of Industrial Engineering, Pusan National University, Busan 609-735, Korea vinhise@pusan.ac.kr Abstract. Accurate demand

More information

Forecasting Geographic Data Michael Leonard and Renee Samy, SAS Institute Inc. Cary, NC, USA

Forecasting Geographic Data Michael Leonard and Renee Samy, SAS Institute Inc. Cary, NC, USA Forecasting Geographic Data Michael Leonard and Renee Samy, SAS Institute Inc. Cary, NC, USA Abstract Virtually all businesses collect and use data that are associated with geographic locations, whether

More information

Improving Students' Data Analysis and Presentation Skills: The Ocean State Circuits, Inc. Forecasting Project

Improving Students' Data Analysis and Presentation Skills: The Ocean State Circuits, Inc. Forecasting Project University of Rhode Island DigitalCommons@URI College of Business Administration Faculty Publications College of Business Administration 2013 Improving Students' Data Analysis and Presentation Skills:

More information

Forecasting sales and intervention analysis of durable products in the Greek market. Empirical evidence from the new car retail sector.

Forecasting sales and intervention analysis of durable products in the Greek market. Empirical evidence from the new car retail sector. Forecasting sales and intervention analysis of durable products in the Greek market. Empirical evidence from the new car retail sector. Maria K. Voulgaraki Department of Economics, University of Crete,

More information

Technical Memorandum Number 794. Expressing Inventory Control Policy in the Turnover Curve

Technical Memorandum Number 794. Expressing Inventory Control Policy in the Turnover Curve Technical Memorandum Number 794 Expressing Inventory Control Policy in the Turnover Curve by Ronald Ballou October 2004 Department of Operations Weatherhead School of Management Case Western Reserve University

More information

Chapter 1. Vector autoregressions. 1.1 VARs and the identi cation problem

Chapter 1. Vector autoregressions. 1.1 VARs and the identi cation problem Chapter Vector autoregressions We begin by taking a look at the data of macroeconomics. A way to summarize the dynamics of macroeconomic data is to make use of vector autoregressions. VAR models have become

More information

The Combination Forecasting Model of Auto Sales Based on Seasonal Index and RBF Neural Network

The Combination Forecasting Model of Auto Sales Based on Seasonal Index and RBF Neural Network , pp.67-76 http://dx.doi.org/10.14257/ijdta.2016.9.1.06 The Combination Forecasting Model of Auto Sales Based on Seasonal Index and RBF Neural Network Lihua Yang and Baolin Li* School of Economics and

More information

A three point formula for finding roots of equations by the method of least squares

A three point formula for finding roots of equations by the method of least squares A three point formula for finding roots of equations by the method of least squares Ababu Teklemariam Tiruneh 1 ; William N. Ndlela 1 ; Stanley J. Nkambule 1 1 Lecturer, Department of Environmental Health

More information

A Seasoned Instructor s New Look At Fixed-Income Securities: An Exercise Using Discounted Bonds Thomas W. Secrest, Coastal Carolina University, USA

A Seasoned Instructor s New Look At Fixed-Income Securities: An Exercise Using Discounted Bonds Thomas W. Secrest, Coastal Carolina University, USA A Seasoned Instructor s New Look At Fixed-Income Securities: An Exercise Using Discounted Bonds Thomas W. Secrest, Coastal Carolina University, USA ABSTRACT Having returned to teaching the basics of pricing

More information

1 Teaching notes on GMM 1.

1 Teaching notes on GMM 1. Bent E. Sørensen January 23, 2007 1 Teaching notes on GMM 1. Generalized Method of Moment (GMM) estimation is one of two developments in econometrics in the 80ies that revolutionized empirical work in

More information

c 2008 Je rey A. Miron We have described the constraints that a consumer faces, i.e., discussed the budget constraint.

c 2008 Je rey A. Miron We have described the constraints that a consumer faces, i.e., discussed the budget constraint. Lecture 2b: Utility c 2008 Je rey A. Miron Outline: 1. Introduction 2. Utility: A De nition 3. Monotonic Transformations 4. Cardinal Utility 5. Constructing a Utility Function 6. Examples of Utility Functions

More information

Chapter 4 Multi-Stage Interconnection Networks The general concept of the multi-stage interconnection network, together with its routing properties, have been used in the preceding chapter to describe

More information

business statistics using Excel OXFORD UNIVERSITY PRESS Glyn Davis & Branko Pecar

business statistics using Excel OXFORD UNIVERSITY PRESS Glyn Davis & Branko Pecar business statistics using Excel Glyn Davis & Branko Pecar OXFORD UNIVERSITY PRESS Detailed contents Introduction to Microsoft Excel 2003 Overview Learning Objectives 1.1 Introduction to Microsoft Excel

More information

American Journal Of Business Education September/October 2012 Volume 5, Number 5

American Journal Of Business Education September/October 2012 Volume 5, Number 5 Teaching Learning Curves In An Undergraduate Economics Or Operations Management Course Jaideep T. Naidu, Philadelphia University, USA John F. Sanford, Philadelphia University, USA ABSTRACT Learning Curves

More information

Oscillations of the Sending Window in Compound TCP

Oscillations of the Sending Window in Compound TCP Oscillations of the Sending Window in Compound TCP Alberto Blanc 1, Denis Collange 1, and Konstantin Avrachenkov 2 1 Orange Labs, 905 rue Albert Einstein, 06921 Sophia Antipolis, France 2 I.N.R.I.A. 2004

More information

Simple formulas to option pricing and hedging in the Black Scholes model

Simple formulas to option pricing and hedging in the Black Scholes model Simple formulas to option pricing and hedging in the Black Scholes model Paolo Pianca Department of Applied Mathematics University Ca Foscari of Venice Dorsoduro 385/E, 3013 Venice, Italy pianca@unive.it

More information

Adaptive model for thermal demand forecast in residential buildings

Adaptive model for thermal demand forecast in residential buildings Adaptive model for thermal demand forecast in residential buildings Harb, Hassan 1 ; Schütz, Thomas 2 ; Streblow, Rita 3 ; Müller, Dirk 4 1 RWTH Aachen University, E.ON Energy Research Center, Institute

More information

Short-cut design of small hydroelectric plants

Short-cut design of small hydroelectric plants Renewable Energy 19 (2000) 545±563 www.elsevier.com/locate/renene Short-cut design of small hydroelectric plants N.G. Voros*, C.T. Kiranoudis, Z.B. Maroulis Department of Chemical Engineering, National

More information

QMB 3302 Business Analytics CRN 10251 Spring 2015 T R -- 11:00am - 12:15pm -- Lutgert Hall 2209

QMB 3302 Business Analytics CRN 10251 Spring 2015 T R -- 11:00am - 12:15pm -- Lutgert Hall 2209 QMB 3302 Business Analytics CRN 10251 Spring 2015 T R -- 11:00am - 12:15pm -- Lutgert Hall 2209 Elias T. Kirche, Ph.D. Associate Professor Department of Information Systems and Operations Management Lutgert

More information

Introduction to time series analysis

Introduction to time series analysis Introduction to time series analysis Margherita Gerolimetto November 3, 2010 1 What is a time series? A time series is a collection of observations ordered following a parameter that for us is time. Examples

More information

Improving Demand Forecasting

Improving Demand Forecasting Improving Demand Forecasting 2 nd July 2013 John Tansley - CACI Overview The ideal forecasting process: Efficiency, transparency, accuracy Managing and understanding uncertainty: Limits to forecast accuracy,

More information