Time Series Analysis and Forecasting (eBook)

Selected Contributions from the ITISE Conference

Ignacio Rojas, Héctor Pomares (Herausgeber)

eBook Download: PDF

2016 | 1st ed. 2016
XIX, 384 Seiten
Springer International Publishing (Verlag)
978-3-319-28725-6 (ISBN)

Infos

This volume presents selected peer-reviewed contributions from The International Work-Conference on Time Series, ITISE 2015, held in Granada, Spain, July 1-3, 2015. It discusses topics in time series analysis and forecasting, advanced methods and online learning in time series, high-dimensional and complex/big data time series as well as forecasting in real problems.

The International Work-Conferences on Time Series (ITISE) provide a forum for scientists, engineers, educators and students to discuss the latest ideas and implementations in the foundations, theory, models and applications in the field of time series analysis and forecasting.  It focuses on interdisciplinary and multidisciplinary research encompassing the disciplines of computer science, mathematics, statistics and econometrics.

Ignacio Rojas is a full professor at the Department of Computer Architecture and Computer Technology and Director of the  Information and Communications Technology Centre (CITIC-UGR), University of Granada, Spain. Throughout his research career, he has served as a principal investigator or participated in more than 20 research projects obtained in competitive calls including projects of the European Union, the I+D+I Spanish National Government and projects Excellence of the Ministry of Innovation, Science and Enterprise Junta de Andalucía. He has published more than 204 scientific contributions reflected in the database ISI Web of Science, thereof 85 articles in JCR-indexed journals.

Héctor Pomares has been a full professor at the University of Granada in Spain since 2001. He has published more than 50 articles in JCR-indexed journals and contributed with more than 150 papers in international conferences. He has led or participated in 15 national projects, one autonomic R&D Excellence project and 13 contracts signed for innovative research through the University of Granada Foundation Company and the Office of Transfer of Research Results. He is a member of the editorial board of the Journal of Applied Mathematics (JCR-indexed) and is the coordinator of the Official Master's Degree in Computer & Network Engineering at the University of Granada.

Inhalt

Preface 6
Contents 11
Contributors 14
Part I Advanced Analysis and Forecasting Methods 19
A Direct Method for the Langevin-Analysis of MultidimensionalStochastic Processes with Strong Correlated Measurement Noise 20
1 Introduction 21
2 Methodology 23
3 Results 25
References 27
Threshold Autoregressive Models for Directional Time Series 29
1 Introduction 29
2 Detecting Directionality 30
2.1 Difference in Linear Quadratic Lagged Correlations 31
2.2 Methods Based on First Differences 32
2.2.1 Percentage of Positive Differences 32
2.2.2 Product Moment Skewness of Differences 32
2.3 Threshold-Peak or Threshold-Trough Test 33
2.4 Evidence of Directionality 33
3 Modeling Directionality 33
4 The Sunspot Series 35
4.1 Directionality in Sunspots 35
4.2 Threshold Autoregressive Model Fitted by Least Squares 36
4.3 Threshold Autoregressive Model Fitted by Penalized Least Squares 36
4.4 Details of Fitting AR(2), TAR(2)[LS], and TAR(2)[LSP] 37
4.5 Simulation to Validate TAR(2)[LS] and TAR(2)[LSP] 38
4.6 Comparisons of One-Step-Ahead Predictions 38
4.7 Comparisons of Distributions of 15-Year Extreme Values 39
5 Conclusion 40
References 41
Simultaneous Statistical Inference in Dynamic Factor Models 42
1 Introduction and Motivation 42
2 Multiple Testing 45
2.1 Multiple Testing Under Positive Dependence 46
2.2 Multivariate Chi-Square Distributed Test Statistics 47
3 Multiple Testing in DFMs 50
4 Finite-Sample Bootstrap Approximation 54
5 Concluding Remarks and Outlook 57
References 58
The Relationship Between the Beveridge–Nelson Decomposition and Exponential Smoothing 61
1 Introduction 61
2 Two Parallel Decompositions of an ARIMA Model 62
2.1 Polynomial Methods 63
2.2 State Space Methods 67
2.3 Connection with Exponential Smoothing 71
References 74
Permutation Entropy and Order Patterns in Long Time Series 75
1 Introduction 75
2 Ups and Downs 77
3 Patterns of Length 3 79
4 Persistence Versus Autocorrelation 82
5 Sliding Windows 83
6 Permutation Entropy and Distance to White Noise 84
7 Partition of the Distance to White Noise 85
8 Biomedical Data 85
References 87
Generative Exponential Smoothing and Generative ARMA Models to Forecast Time-Variant Rates or Probabilities 88
1 Introduction 88
2 Related Work 89
3 Generative Models for Rates or Probability Estimates 90
3.1 Basic Idea 90
3.2 Generative Exponential Smoothing 92
3.3 Generative Exponential Smoothing with Trend 93
3.4 Generative Double Exponential Smoothing 93
3.5 Generative Autoregressive Model 94
3.6 Generative Autoregressive Moving Average 96
4 Simulation Results 97
4.1 Data Sets 97
4.2 Forecasting Performance 98
4.3 Run-Time 99
5 Conclusion and Outlook 100
References 101
First-Passage Time Properties of Correlated Time Series with Scale-Invariant Behavior and with Crossovers in the Scaling 102
1 Introduction 103
2 FPT Characteristics of Time Series with Scale-Invariant Behavior 104
3 Algorithm Generating Time Series with Crossovers in the Scaling 106
4 FPT Probability Density for Time Series with Crossovers in the Scaling 107
5 Mean FPT Value for Time Series with Crossovers in the Scaling 110
6 Comparing FPT Theoretical Predictions with Experimental Observations for Systems with Crossovers in the Scaling 111
7 Conclusions 113
Appendix 113
References 114
Part II Theoretical and Applied Econometrics 116
The Environmental Impact of Economic Activity on the Planet 117
1 Introduction and Objectives 117
2 Methodology 118
3 Results 120
4 Conclusions 123
References 123
Stock Indices in Emerging and Consolidated Economies from a Fractal Perspective 125
1 Introduction 126
2 Exponent of Colored Noise 127
3 Fractional Brownian Motion 129
3.1 Statistical Tests 132
4 Conclusions 132
References 134
Value at Risk with Filtered Historical Simulation 135
1 Introduction 135
2 An Overview of the Historical Simulation VaR 136
3 Empirical Investigation 139
3.1 Descriptive Statistics 139
4 Conclusion 144
References 144
A SVEC Model to Forecast and Perform Structural Analysis (Shocks) for the Mexican Economy, 1985Q1 –2014Q4 146
1 Introduction 146
2 Literature Review 147
2.1 Justification and Variable Selection 147
2.1.1 Monetary Policy (M2) 147
2.1.2 Unemployment 148
2.1.3 GDP and Real Exchange Rate 148
2.1.4 US Industrial Output and Economic Integration 148
3 Stylized Facts 148
4 Econometric Issues 149
4.1 Estimation 149
4.2 Structural Analysis: Methodology 151
4.3 Identification and Analysis of Results 152
4.3.1 Monetary Effects and Okun's Law ( Z1=ut,yt,m2t ) 152
4.3.2 External Sector ( Z2=yust,yt,qt ) 153
5 Conclusion 154
6 Statistical Appendix 155
References 155
IntradayDatavsDailyDatatoForecastVolatilityinFinancialMarkets 157
1 Introduction 157
2 Realized Volatility and Integrated Volatility 158
3 The Stochastic Volatility Model 159
3.1 Stochastic Volatility Estimation 160
4 Particle Filter Methods 161
4.1 Particle Filter for the SV Model 162
5 An Empirical Demonstration 163
5.1 Data Description 163
5.2 Parameters Estimation for the SV Model 164
5.3 Volatility Measures Comparison 166
6 Conclusion 167
References 168
Predictive and Descriptive Qualities of Different Classes of Models for Parallel Economic Development of Selected EU-Countries 170
1 Introduction 171
2 Theoretical Background 173
2.1 Aggregation Functions 173
2.2 Regime-Switching Models 175
2.2.1 TAR Models 175
2.2.2 STAR Models 176
2.2.3 Markov-Switching Models MSW 178
2.3 Model Specification 178
3 Modelling Results 180
4 Conclusions 181
References 181
Search and Evaluation of Stock Ranking Rules Using Internet Activity Time Series and Multiobjective Genetic Programming 183
1 Introduction 183
2 Related Research 184
3 Goal 186
4 Methods 186
5 Results 188
6 Conclusion 193
References 193
Integer-Valued APARCH Processes 196
1 Introduction 196
2 Integer-Valued APARCH(p,q) Processes 197
3 Parameter Estimation 201
4 Simulation 204
4.1 Log-Likelihood Analysis 206
5 Real-Data Example: Transaction Modeling 207
References 208
Part III Applications in Time Series Analysis and Forecasting 210
Emergency-Related, Social Network Time Series: Descriptionand Analysis 211
1 Introduction 211
2 Incentives for and Limits in the Use of Analytics in Disasters 213
3 SN Response and Related Time Series 214
4 Examples of SN-Related Time Series 216
5 Discussion and Conclusions 218
References 219
Competitive Models for the Spanish Short-Term Electricity Demand Forecasting 222
1 Introduction 222
2 Spanish Short-Term Electricity Demand Forecasting 225
2.1 REE Modelling 225
3 Multiple Seasonal Exponential Smoothing Holt–Winters Models 226
3.1 Holt–Winters Models Generalisation 226
3.2 Computational Strategy for Optimising the Smoothing Parameters 230
4 Modelling Approach and Results 230
4.1 Model Selection 231
4.2 Validation 232
5 Conclusions 234
References 235
Age-Specific Death Rates Smoothed by the Gompertz–Makeham Function and Their Application in Projections by Lee–Carter Model 237
1 Introduction and Literature Review 237
2 Materials and Used Methods 238
3 Results and Discussion 244
4 Conclusion 248
References 248
An Application of Time Series Analysis in Judging the Working State of Ground-Based Microwave Radiometers and Data Calibration 250
1 Introduction 251
2 Clear-Sky Brightness Temperature Simulations and Observations 252
3 Comparison Between the Simulated and the Observed Brightness Temperature Series 253
3.1 At the Frequencies for Water Vapor and Liquid Water Remote Sensing 253
3.2 At the Frequencies for Air Temperature Remote Sensing 255
4 Applications of the Method onto Other Radiometers 257
4.1 The Situation of the Radiometers at Wuhan 257
4.2 The Situation of the Beijing Radiometer 257
5 Conclusion 260
References 261
Identifying the Best Performing Time Series Analytics for Sea Level Research 263
1 Introduction 263
2 Method 264
2.1 Step 1: Development of Synthetic Data Sets for Testing Purposes 265
2.2 Step 2: Application of Analysis Methods to Extract Trend from Synthetic Data Sets 265
2.3 Step 3: Multi-Criteria Assessment of Analytical Methods for Isolating Mean Sea Level 268
3 Results 269
4 Discussion 273
5 Conclusion 276
References 277
Modellation and Forecast of Traffic Series by a Stochastic Process 281
1 Introduction 281
2 State of the Art 284
3 Methodology 284
3.1 Preprocessing Data 285
3.1.1 Week-Long Paths: Series A 285
3.1.2 Day-Long Paths: Series B, C and D 287
3.2 The Gompertz-Lognormal Diffusion Process 287
4 Results 290
5 Conclusions 293
References 294
Spatio-Temporal Modeling for fMRI Data 295
1 Introduction 295
2 Method 297
2.1 Model 298
2.2 OLS Estimate 299
3 Hypothesis Testing 299
3.1 Key Concepts 299
3.2 Testing the Linearity 301
3.3 Testing the Effect from a Specific Stimulus 301
3.4 Detecting the Activation 302
3.5 Testing the Difference Between HRFs 302
3.6 Remarks 303
4 Simulation 305
5 Real Data Analysis 309
5.1 Auditory Data 309
6 Discussion 311
References 312
Part IV Machine Learning Techniques in Time Series Analysis and Prediction 314
Communicating Artificial Neural Networks with Physical-Based Flow Model for Complex Coastal Systems 315
1 Introduction 316
2 Brief Description of the Study Site 316
3 Knowledge Recovery System Using ANNs 317
4 Existing Knowledge Base During 2004–2005 Periods 318
5 Developments of DRS for Inland Boundary Conditions During 2004–2005 Periods 320
6 Development of Inland Flow and Salinity Boundary Simulators (WMC-GIWW) During 2004–2005 Periods 321
7 Patterns Search from the Historical Hydrological Record 322
8 Inland Flow and Salinity Boundary Conditions Simulation During Selected Average Flow Year, High and Low Flow Months 323
9 Learning Remarks 325
10 Conclusions 326
References 327
Forecasting Daily Water Demand Using Fuzzy Cognitive Maps 328
1 Introduction 328
2 Theoretical Background 329
2.1 Forecasting Time Series 330
2.2 FCM Fundamentals 331
3 Mapping FCM Model to Water Demand Time Series 332
4 Parametrized Membership Function 334
5 Experiments 334
5.1 Quality of Data 334
5.2 Statistical Features of Data 335
5.3 Experimental Setup 336
6 Conclusion 337
References 337
Forecasting Short-Term Demand for Electronic Assemblies by Using Soft-Rules 340
1 Introduction 341
2 Learning Soft-Rules 342
2.1 Data Normalization and the Study-Pairs 342
2.2 Forming Soft-Rules 343
3 Using Soft-Rules for Demand Forecasting 349
4 Empirical Results and Conclusions 350
References 352
Electrical Load Forecasting: A Parallel Seasonal Approach 353
1 Introduction 353
2 Related Work 354
3 Models Used 355
3.1 Multivariable Regression 355
3.2 Artificial Neural Network 356
3.2.1 Multi-Layer Perceptron 356
3.3 Support Vector Machines 356
3.3.1 Support Vector Regression 357
4 Methodology 357
4.1 Data Selection and Preprocessing 359
4.2 Considered Variables 360
4.3 Design of the Proposed SVR and MLP ANN Models 360
4.4 From the Long to Short Term 360
5 Results and Discussion 361
6 Conclusion 363
References 363
A Compounded Multi-resolution-Artificial Neural Network Method for the Prediction of Time Series with Complex Dynamics 365
1 Introduction 365
1.1 Signal Decomposing and Prediction Procedures 366
1.1.1 The Learning Algorithm and the Regularization Parameter 368
1.1.2 Intelligent Network Parameters Optimization 368
1.1.3 Human-Driven Decisions 370
1.1.4 The Algorithm 370
2 Empirical Analysis 373
2.1 Results 376
References 382
Erratum to: Stock Indices in Emerging and Consolidated Economies from a Fractal Perspective 383

Produkt-Details

• Erscheint lt. Verlag: 30.5.2016
• Reihe/Serie: Contributions to Statistics
• Reihe/Serie: Contributions to Statistics
• Zusatzinfo: XIX, 384 p. 112 illus., 49 illus. in color.
• Verlagsort: Cham
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik
• Themenwelt: Mathematik / Informatik ► Mathematik ► Statistik
• Themenwelt: Mathematik / Informatik ► Mathematik ► Wahrscheinlichkeit / Kombinatorik
• Themenwelt: Technik
• Themenwelt: Wirtschaft ► Allgemeines / Lexika
• Schlagworte: 37M10, 62M10, 62-XX, 68-XX, 60-XX, 58-XX, 37-XX • applications in computer science • applications in econometrics • applications in industry • Big Data • Forecasting • high-dimensional data • machine learning • on-line learning • real-life problems • statistical methods for time series • Time Series Analysis
• ISBN-10: 3-319-28725-7 / 3319287257
• ISBN-13: 978-3-319-28725-6 / 9783319287256