Buch, Englisch, 464 Seiten, Format (B × H): 203 mm x 254 mm, Gewicht: 1157 g
ISBN: 978-1-119-58550-3
Verlag: Wiley
Brings mathematics to bear on your real-world, scientific problems
Mathematical Methods in Interdisciplinary Sciences provides a practical and usable framework for bringing a mathematical approach to modelling real-life scientific and technological problems. The collection of chapters Dr. Snehashish Chakraverty has provided describe in detail how to bring mathematics, statistics, and computational methods to the fore to solve even the most stubborn problems involving the intersection of multiple fields of study. Graduate students, postgraduate students, researchers, and professors will all benefit significantly from the author's clear approach to applied mathematics.
The book covers a wide range of interdisciplinary topics in which mathematics can be brought to bear on challenging problems requiring creative solutions. Subjects include:
- Structural static and vibration problems
- Heat conduction and diffusion problems
- Fluid dynamics problems
The book also covers topics as diverse as soft computing and machine intelligence. It concludes with examinations of various fields of application, like infectious diseases, autonomous car and monotone inclusion problems.
Autoren/Hrsg.
Fachgebiete
- Mathematik | Informatik Mathematik Numerik und Wissenschaftliches Rechnen Angewandte Mathematik, Mathematische Modelle
- Mathematik | Informatik Mathematik Numerik und Wissenschaftliches Rechnen Mathematische Modellierung
- Technische Wissenschaften Maschinenbau | Werkstoffkunde Maschinenbau Konstruktionslehre, Bauelemente, CAD
- Technische Wissenschaften Technik Allgemein Computeranwendungen in der Technik
- Technische Wissenschaften Maschinenbau | Werkstoffkunde Technische Mechanik | Werkstoffkunde Werkstoffkunde, Materialwissenschaft: Forschungsmethoden
Weitere Infos & Material
Notes on Contributors xv
Preface xxv
Acknowledgments xxvii
1 Connectionist Learning Models for Application Problems Involving Differential and Integral Equations 1
Susmita Mall, Sumit Kumar Jeswal, and Snehashish Chakraverty
1.1 Introduction 1
1.1.1 Artificial Neural Network 1
1.1.2 Types of Neural Networks 1
1.1.3 Learning in Neural Network 2
1.1.4 Activation Function 2
1.1.4.1 Sigmoidal Function 3
1.1.5 Advantages of Neural Network 3
1.1.6 Functional Link Artificial Neural Network (FLANN) 3
1.1.7 Differential Equations (DEs) 4
1.1.8 Integral Equation 5
1.1.8.1 Fredholm Integral Equation of First Kind 5
1.1.8.2 Fredholm Integral Equation of Second Kind 5
1.1.8.3 Volterra Integral Equation of First Kind 5
1.1.8.4 Volterra Integral Equation of Second Kind 5
1.1.8.5 Linear Fredholm Integral Equation System of Second Kind 6
1.2 Methodology for Differential Equations 6
1.2.1 FLANN-Based General Formulation of Differential Equations 6
1.2.1.1 Second-Order Initial Value Problem 6
1.2.1.2 Second-Order Boundary Value Problem 7
1.2.2 Proposed Laguerre Neural Network (LgNN) for Differential Equations 7
1.2.2.1 Architecture of Single-Layer LgNN Model 7
1.2.2.2 Training Algorithm of Laguerre Neural Network (LgNN) 8
1.2.2.3 Gradient Computation of LgNN 9
1.3 Methodology for Solving a System of Fredholm Integral Equations of Second Kind 9
1.3.1 Algorithm 10
1.4 Numerical Examples and Discussion 11
1.4.1 Differential Equations and Applications 11
1.4.2 Integral Equations 16
1.5 Conclusion 20
References 20
2 Deep Learning in Population Genetics: Prediction and Explanation of Selection of a Population 23
Romila Ghosh and Satyakama Paul
2.1 Introduction 23
2.2 Literature Review 23
2.3 Dataset Description 25
2.3.1 Selection and Its Importance 25
2.4 Objective 26
2.5 Relevant Theory, Results, and Discussions 27
2.5.1 automl 27
2.5.2 Hypertuning the Best Model 28
2.6 Conclusion 30
References 30
3 A Survey of Classification Techniques in Speech Emotion Recognition 33
Tanmoy Roy, Tshilidzi Marwala, and Snehashish Chakraverty
3.1 Introduction 33
3.2 Emotional Speech Databases 33
3.3 SER Features 34
3.4 Classification Techniques 35
3.4.1 Hidden Markov Model 36
3.4.1.1 Difficulties in Using HMM for SER 37
3.4.2 Gaussian Mixture Model 37
3.4.2.1 Difficulties in Using GMM for SER 38
3.4.3 Support Vector Machine 38
3.4.3.1 Difficulties with SVM 39
3.4.4 Deep Learning 39
3.4.4.1 Drawbacks of Using Deep Learning for SER 41
3.5 Difficulties in SER Studies 41
3.6 Conclusion 41
References 42
4 Mathematical Methods in Deep Learning 49
Srinivasa Manikant Upadhyayula and Kannan Venkataramanan
4.1 Deep Learning Using Neural Networks 49
4.2 Introduction to Neural Networks 49
4.2.1 Artificial Neural Network (ANN) 50
4.2.1.1 Activation Function 52
4.2.1.2 Logistic Sigmoid Activation Function 52
4.2.1.3 tanh or Hyperbolic Tangent Activation Function 53
4.2.1.4 ReLU (Rectified Linear Unit) Activation Function 54
4.3 Other Activation Functions (Variant Forms of ReLU) 55
4.3.1 Smooth ReLU 55
4.3.2 Noisy ReLU 55
4.3.3 Leaky ReLU 55
4.3.4 Parametric ReLU 56
4.3.5 Training and Optimizing a Neural Network Model 56
4.4 Backpropagation Algorithm 56
4.5 Performance and Accuracy 59
4.6 Results and Observation 59
References 61
5 Multimodal Data Representation and Processing Based on Algebraic System of Aggregates 63
Yevgeniya Sulema and Etienne Kerre
5.1 Introduction 63
5.2 Basic Statements of ASA 64
5.3 O
