旅游预测方法手册

A report produced forthe European Travel Commissionby Inzights & Silverbullet Research .........................................................................9..............................................................................9 ..............................................15.................................................................16............................................................24.............................................26.......................................................................................................................28...........................32..32..........................................35........................................37....................................................................................................................39....................................................................................46..................................................................................................................................48....................................49.................................................................................51..55 .......................59.............................................................................................62..................................................................63.................................................................................64................................................................................65.................................................................................69 ......................................................................................78 ...89......................................................................................................................92 1.1 THE RELEVANCE OF FORECASTING 1.2WHAT TO FORECAST–AND WHY VARIABLES DISTINCTIONS AND IMPLICATIONS OF LONG- AND SHORT-TERM FORECASTING STRUCTURE FOR THE BASIC DESCRIPTION OF FORECASTING METHODS AND INTRODUCTION OF THECASE DATA SET 2.1TOURISM FORECASTING WITH CASUAL MODELS 2.1.1 SIMPLE LINEAR REGRESSION CASE EXAMPLE: FORECASTING BED NIGHTS WITH SIMPLE LINEAR REGRESSION STEP 1: FORECASTING BED NIGHTS IN COPENHAGEN IN 2019 WITH SIMPLE LINEAR REGRESSION STEP 2: FORECASTING BED NIGHTS IN COPENHAGEN IN 2019 WITH TWO-RAPAMETER CURVE FITTING STEP 3: FORECASTING BED NIGHTS IN COPENHAGEN IN 2019 WITH A MODIFIED SIMPLE LINEAR REGRESSIONUSING ONLY OBSERVATIONS SINCE 2010 2.1.2 MULTIPLE LINEAR REGRESSION CASE EXAMPLE: FORECASTING BED NIGHTS WITH MULTIPLE LINEAR REGRESSION 2.1.3 STRUCTURAL ECONOMETRIC METHODS 2.1.4 ADVANTAGES AND DISADVANTAGESOF THE LINEAR REGRESSION MODEL STEP-BY-STEP GUIDE–LINEAR REGRESSION SUGGESTIONS FOR FURTHER READING BIBLIOGRAPHY 2.2 TOURISM FORECASTING WITH EXTRAPOLATIVE METHODS 2.2.1 NO-CHANGE METHODS AND SIMPLE MOVING AVERAGE (SMA) CASE EXAMPLE: FORECASTING BED NIGHTS WITH NAÏVE MODELS AND SMA STEP-BY-STEP GUIDE–SIMPLE MOVING AVERAGE 2.2.2 SIMPLE EXPONENTIAL SMOOTHING (SES) CASE EXAMPLE: FORECASTING BED NIGHTS WITH SES STEP-BY-STEP GUIDE–SIMPLE EXPONENTIAL SMOOTHING 2.2.3 DOUBLE EXPONENTIAL SMOOTHING (DES) CASE EXAMPLE: FORECASTING BED NIGHTS WITH HES STEP-BY-STEP GUIDE–DOUBLE EXPONENTIAL SMOOTHING 2.2.4 AUTOREGRESSIVE MOVING AVERAGE (ARMA), ARIMA AND SARIMA MODELS CASE EXAMPLE: FORECASTING BED NIGHTS WITH SARIMA STEP-BY-STEP GUIDE STEP 1: EXPLORATIVE DATA ANALYSIS STEP 2: IDENTIFICATION OF THE MOST SUITABLE SARIMA MODEL TYPE •• STEP 3: ESTIMATING THE PARAMETERS FOR THE CHOSEN MODEL STEP 4: MAKE FORECASTS STEP-BY-STEP GUIDE–SARIMA/ARIMA MODELLING DISCUSSION OF RESULTS 2.2.5 ADVANTAGES AND DISADVANTAGES OF EXTRAPOLATIVE FORECASTING METHODS SUGGESTIONS FOR FURTHER READING 2.3 TOURISM FORECASTING WITH ARTIFICIAL INTELLIGENCE AND HYBRID MODELS 2.3.1 ARTICIFIAL INTELLIGENCE FUNDAMENTALS SUPERVISED MACHINE LEARNING ϵϵϵ ϵ UNSUPERVISED MACHINE LEARNING BLACK BOX IS A CHALLENGE AI METHODOLOGY IS EVOLVING 2.3.2 A BRIEF OVERVIEW OF METHODS AND TERMINOLOGY IN ARTIFICIAL INTELLIGENCE ARTIFICIAL NEUTRAL NETWORKS (ANN) SUPPORT VECTOR MACHINES (SVM) DEEP LEARNING EXTREME LEARNING MACHINE (ELM) MULTILAYER PERCEPTRON REGRESSION (MLP) FUZZY TIME SERIES A GREY SYSTEM SWARM INTELLIGENCE SOFT INTELLIGENCE HYBRID MODELS 2.3.3 EXAMPLES OF APPLICATION OF AI IN TOURISM FORECASTING CASE A: USING ARTIFICIAL INTELLIGENCE TO FORECAST HOTEL CAPACITY IN COPENHAGEN CASE B: FORECASTING TOURISM IN GREESE USING NEUTRAL NETWORKS CASE C: MULTIVARIATE FORECASTING MODEL FOR TOURISM IN THE LISBON AREA CASE D: PARTICLE SWARM OPTIMISATION SVR FOR TOURISTS TO TAIWAN CASE E: HYBRID DEEP LEARNING ON TOURISTS IN SINGAPORE ffff CASE F: USING BIG DATA TO PREDICT TOURIST ARRIVALS IN BEIJING AND VIENNA 2.3.4 ADVANTAGES AND DISADVANTAGES OF AI MODELS STEP-BY-STEP GUIDE–CHOOSING AN AI MODEL FOR TOURISM FORECASTING SUGGESTIONS FOR FURTHER READING BIBLIOGRAPHY 2.4 QUALITATIVE METHODS 2.