Operations Research: New Paradigms and Emerging Applications

Cover
Title Page
Copyright Page
Foreword
Preface
Acknowledgement
Table of Contents
Contributors
Part I: Operation Research in Optimization
1. Kernel Based C-Bridge Estimator for Partially Nonlinear Model
1.1 Introduction
1.2 Bridge Estimators
1.3 PNLMs with Additive Approximation and Bridge Estimation
1.3.1 Construction of additive nonparametric component
1.3.2 Kernel based bridge estimation for PNLMs
1.4 On Conic Optimization and Its Application to Kernel Based Bridge Problem
1.4.1 Convex and conic optimization
1.4.2 Conic kernel based bridge estimator (C-KBBE)
1.4.2.1 C-KBBE for case (α = 2)
1.4.2.2 C-KBBE for case (α = 1)
1.5 Conclusion
References
2. A Glimpse on the Contributions and Challenges towards more Environmentally-friendly Road Traffic
2.1 Introduction
2.2 Research Methodology and Descriptive Analysis
2.3 Exploring CAVs Environmental Impacts
2.3.1 Eco-routing and control strategies
2.3.2 Model adjustment framework
2.4 Conclusions and Future Research Directions
References
3. ELECTRE I for Balancing Projects: Case Studies for Selecting Suppliers and Portfolio Investment Schemes
3.1 Introduction
3.2 Milestones in the Family of ELECTRE Methods
3.3 Context of Application of the ELECTRE Methods
3.4 Theoretical and Conceptual Explanation of ELECTRE I
3.4.1 Decision matrix
3.4.2 Matrix of concordance indices
3.4.3 Normalized decision matrix (NMD)
3.4.4 Normalized and weighted decision matrix (NMDP)
3.4.5 Matrix of disagreement indices
3.4.6 Concordance threshold (c*) and discordance threshold (d*)
3.4.7 Concordant dominance matrix (CDM) and discordant dominance matrix (DDM)
3.4.8 Aggregate dominance matrix (ADM)
3.4.9 ELECTRE graph
3.5 Application Examples
3.6 Conclusions
References
Part II: Operation Research in Data Mining and Clustering
4. Semidefinite Optimization Models in Multivariate Statistics: A Survey
4.1 Introduction
4.2 Semidefinite Programming (SDP)
4.2.1 SDP problems: basic definitions and duality results
4.2.2 SDP solvers
4.3 SDP Application to Multivariate Statistical Techniques
4.3.1 Principal component analysis
4.3.2 Clustering
4.3.3 Clustering and disjoint principal component analysis
4.4 Conclusions
References
5. Operation Research Techniques in Data Mining Focusing on Clustering
5.1 Introduction
5.2 Background on Clustering and Main Clustering Methods in Data Mining
5.2.1 Measures of similarity and dissimilarity
5.2.2 Clustering methods in data mining
5.2.2.1 Hierarchical clustering
5.2.2.2 Partitioning clustering
5.2.2.3 Density based clustering
5.2.2.4 Grid based clustering
5.3 Formulation of Clustering Problems as an Optimization Problem
5.4 Operations Research Applications of Clustering Algorithms
5.5 Conclusion and Outlook
References
6. Data Mining Approaches to Meteorological Data: A Review of NINLIL Climate Research Group Studies
6.1 Introduction
6.2 Knowledge Discovery in Databases and Data Mining
6.2.1 Data preparation and preprocessing
6.2.2 Data mining
6.2.3 Evaluation, interpretation and implementation
6.3 Climate Data Preparation and Preprocessing
6.3.1 Check for homogeneity of data
6.3.2 Missing data handling
6.4 Climate Data Mining
6.4.1 Determine if the climate has changed by descriptive mining summarization
6.4.2 Clustering climate regions
6.4.3 Identifying seasons by clustering
6.4.4 Precipitation modeling
6.5 Conclusions and Future Studies
References
Part III: Operation Research in Business Science and Finance
7. Fundamentals of Market Making Via Stochastic Optimal Control
7.1 Introduction
7.2 Model Dynamics
7.3 Optimal Quotes under Inventory Risk and Market Impact
7.4 A Focus: Short-term-alpha
7.5 Market Making in an Options Market
7.6 Conclusion and Outlook
References
8. General Points of the Multi-criteria Flow Problems
8.1 Introduction
8.2 Basic Concepts of the Multi-Criteria Decision Aid
8.2.1 Definitions of the set of actions and criteria
8.2.2 Definition of a multi-criteria problem
8.3 Concepts of the Graph’s Theory and Linear Programming
8.3.1 Definitions
8.3.2 Hyperplans and half-spaces
8.3.3 A linear program definition
8.4 Linear Programming Geometry
8.5 Concepts of the Algorithmic Complexity
8.5.1 The flow problems resolution method
8.6 Conclusion
References
9. Operation Research in Neuroscience: A Recent Perspective of Operation Research Application in Finance
9.1 Introduction
9.2 Machine Learning Techniques
9.2.1 Multivariate adaptive regression splines (MARS)
9.2.2 Random forest algorithm
9.2.3 Neural network
9.2.3.1 Single-layer neural network: The perceptron
9.2.3.2 Multilayer neural networks
9.3 Application of Machine Learning Techniques into Investor Sentiment
9.3.1 Human factor: investor sentiment
9.3.2 Application
9.4 Conclusion
References
Part IV: Operation Research in Medical Application
10. An Algorithm and Stability Approach for the Acute Inflammatory Response Dynamic Model
10.1 Introduction
10.2 Mathematical Model
10.3 Numerical Method
10.3.1 Fundamental matrix relations
10.3.2 The collocation approach
10.3.3 Convergence and error bounds
10.3.4 The algorithm
10.4 Stability Analysis
10.4.1 Equilibrium of the model
10.4.2 Linearisation
10.5 Numerical Simulations
10.6 Conclusion and Outlook
References
11. Bayesian Inference for Undirected Network Models
11.1 Introduction
11.2 Copula Gaussian Graphical Model (CGGM)
11.2.0.1 Gaussian copula
11.2.1 Reversible jump Markov chain Monte Carlo method (RJMCMC)
11.2.2 RJMCMC with birth-and-death moves
11.2.3 RJMCMC with split-merge moves
11.3 RJMCMC Alternatives
11.3.1 Birth-and-death MCMC (BDMCMC)
11.3.2 Carlin-Chib algorithm
11.3.3 Gibbs sampling
11.3.4 Quadratic approximation for sparse inverse covariance estimation (QUIC)
11.4 Copula
11.4.1 The Elliptical copulas
11.4.2 The Archimedean copula
11.5 Vine Copula in Inference of Complex Data
11.6 Application
11.7 Discussion
References
12. Evaluation of Data Compression Methods for Efficient Transport and Classification of Facial EMG Signals
12.1 Introduction
12.2 Background
12.2.1 Characteristics of the EMG signal
12.2.2 Compression and classification techniques used in EMG
12.2.3 Applications of EMG
12.2.4 Optimization of cost and performance
12.3 Methods
12.3.1 EMG compression techniques to be implemented
12.3.1.1 Discrete cosine transform
12.3.1.2 Principle component analysis (PCA)
12.3.2 Emotion classification techniques to be implemented
12.3.2.1 Tree classifier
12.3.2.2 K-nearest neighbour (K-NN) classifier
12.3.3 Use case: Prediction of fear from EMG
12.4 Results
12.4.1 Performance
12.4.2 Computational cost
12.5 Discussion and Conclusion
References
Index