Advanced Machine Learning with Evolutionary and Metaheuristic Techniques

✍ Scribed by Jayaraman Valadi, Krishna Pratap Singh, Muneendra Ojha, Patrick Siarry

• Publisher: Springer
• Year: 2024
• Tongue: English
• Leaves: 365
• Category: Library

✦ Synopsis

This book delves into practical implementation of evolutionary and metaheuristic algorithms to advance the capacity of machine learning. The readers can gain insight into the capabilities of data-driven evolutionary optimization in materials mechanics, and optimize your learning algorithms for maximum efficiency. Or unlock the strategies behind hyperparameter optimization to enhance your transfer learning algorithms, yielding remarkable outcomes. Or embark on an illuminating journey through evolutionary techniques designed for constructing deep-learning frameworks. The book also introduces an intelligent RPL attack detection system tailored for IoT networks. Explore a promising avenue of optimization by fusing Particle Swarm Optimization with Reinforcement Learning. It uncovers the indispensable role of metaheuristics in supervised machine learning algorithms. Ultimately, this book bridges the realms of evolutionary dynamic optimization and machine learning, paving the way for pioneering innovations in the field.

✦ Table of Contents

Preface
Contents
From Evolution to Intelligence: Exploring the Synergy of Optimization and Machine Learning
1 Introduction
2 Preliminary Concepts
2.1 Evolutionary Optimization Techniques
2.2 Machine Learning
3 The Synergy of Evolutionary Optimization Techniques and Machine Learning
4 Conclusion
References
Metaheuristic and Evolutionary Algorithms in Explainable Artificial Intelligence
1 Explainable Artificial Intelligence (XAI)
2 Evolutionary and Metaheuristic Algorithms
2.1 Genetic Algorithms
2.2 Particle Swarm Optimization
2.3 Differential Evolution
2.3.1 Types of Vectors
2.4 Multi-objective Optimization Using Non-dominated Sorting Genetic Algorithm II
2.5 Genetic Programming
3 Application of Metaheuristic and Evolutionary Algorithms in XAI
3.1 Counterfactual Explanations
3.1.1 Data Types and Distance Measures
3.1.2 Counterfactuals Using Genetic Algorithms
3.1.3 Counterfactuals Using Particle Swarm Optimization and Differential Evolution
3.1.4 Multi Objective Counterfactuals Using NSGA-II
3.1.5 Additional Works
3.2 Local Surrogate Modelling
3.2.1 Local Rule-Based Explanations
3.2.2 Genetic Programming Explainer
3.3 Transparent Models
3.3.1 Decision Tree
3.3.2 Learning Classifier System
4 Concluding Remarks
References
Evolutionary Dynamic Optimization and Machine Learning
1 Introduction
2 Evolutionary Dynamic Optimization
3 Machine Learning
4 Machine Learning for Resolving Dynamic Optimization Problems
4.1 Transfer Learning-Based
4.2 Supervised Learning-Based
4.3 Reinforcement Learning-Based
4.4 Unsupervised Learning-Based
4.5 Other Learning-Based Models
5 Using Evolutionary Dynamic Optimization in Machine Learning
6 Conclusion
References
Evolutionary Techniques in Making Efficient Deep-Learning Framework: A Review
1 Introduction
2 Theoretical Background
2.1 Deep Learning Models
2.2 Evolutionary Algorithms
2.3 Taxonomy
3 Literature Survey
3.1 Data Preprocessing
3.2 Model Search
3.3 Model Training
3.4 Model Evaluation and Utilization
4 Results and Discussion
5 Conclusion and Future Scope
References
Integrating Particle Swarm Optimization with Reinforcement Learning: A Promising Approach to Optimization
1 Introduction
2 Overview of Methodologies in Literature
2.1 Reinforcement Learning
2.2 Deep Learning Approach for RL
2.3 Particle Swarm Optimization
2.4 Hybrid RL with PSO
3 PSO Implementation for Optimization
3.1 Proposed Model
3.2 Experimental Setup
4 Result and Discussion
5 Conclusion and Future Direction
References
Synergies Between Natural Language Processing and Swarm Intelligence Optimization: A Comprehensive Overview
1 Introduction
2 Natural Language Processing Techniques
2.1 Overview of Natural Language Processing Techniques for Text Analysis
2.2 Applications of Natural Language Processing
2.3 NLP Technological Evolution
2.4 Challenges in NLP Applications
3 Introduction to Swarm Intelligence
3.1 ACO
3.2 PSO
3.3 Artificial Bee Colony
3.4 Firefly (FA)
3.5 Grey Wolf Optimizations (GWO)
4 Combining Natural Language Processing and Swarm Intelligence Optimization
4.1 Pipeline for Swarm Intelligence-Based Optimization Based Natural Language Processing
4.2 Hyperparameter Tuning with Swarm Optimization
4.3 Swarm Intelligence Based Pretrain Language Model
4.4 Swarm Intelligence in Transfer Learning-Based Language Processing
4.5 Swarm Intelligence and Large-Scale Language Model
5 Use Case or Application Where Integration of SI and NLP Can Be Beneficial
5.1 Sentiment Analysis with SI
5.2 Social Media Analytics with SI
5.3 Topic Modeling with SI
6 Challenges and Future Directions
7 Conclusion and Summary
References
Heuristics-Based Hyperparameter Tuning for Transfer Learning Algorithms
1 Introduction
2 Related Works
3 Evolutionary Algorithms
3.1 Categorization of Evolutionary Algorithms
4 Particle Swarm Optimization
4.1 Mathematical Formulation
5 Transfer Learning
5.1 Mathematical Formulation
5.2 What Can Be Transferred?
5.3 Categorization of Transfer Learning Algorithms
5.3.1 On the Basis of Effect
5.3.2 On the Basis of Availability of Label Information for Transfer
5.3.3 Inductive Transfer Learning
5.3.4 Transductive Transfer Learning
5.3.5 Unsupervised Transfer Learning
5.3.6 Multitask Learning
5.3.7 On the Basis of Size of Target Domain
6 Self-Taught Learning
6.1 Mathematical Formulation
6.2 Proposed Methodology
6.3 Experiments
6.4 Results and Discussion
7 Zero-Shot Learning
7.1 Mathematical Formulation
7.2 Side Information: Role and the Different Sources
7.3 Proposed Methodology
7.4 Optimization Process
7.5 Model Testing
7.6 Label Propagation
7.7 Experiments
7.8 Results and Discussion
8 Key Takeaways
8.1 Role of Hyperparameters
8.2 Obtaining Trainables as a Byproduct of Hyperparameter Tuning Process
8.3 Exploration Vs. Exploitation
9 Conclusion
References
Machine Learning Applications of Evolutionary and Metaheuristic Algorithms
1 Introduction
1.1 Advantages/Disadvantages of Evolutionary Algorithms Over Classical Algorithms
2 Single-Objective Optimization Problems in Machine Learning
2.1 Definition
2.2 Formulation
2.3 Particle Swarm Optimization (PSO)
2.3.1 Applications of PSO in Machine Learning
2.4 Ant Colony Optimization (ACO)
2.4.1 Applications of ACO in Machine Learning
2.5 Artificial Bee Colony (ABC)
2.5.1 Applications of ABC in Machine Learning
2.6 Differential Evolution (DE)
2.6.1 Applications of DE in Machine Learning
3 Multi-objective Optimization Problem
3.1 Definition
3.2 Formulation
3.3 Objectives in Multi-objective Optimization
4 Major Methods to Deal with MOOP
4.1 Classical Methods
4.1.1 Weighted Sum Method
4.1.2 - Constraint Method
4.1.3 Weighted Metric Method
5 Applications of Multi-objective Evolutionary Algorithms in Machine Learning
5.1 Nondominated Sorting Genetic Algorithm II (NSGA-II)
5.1.1 Applications of NSGA-II in Machine Learning
5.2 Niched-Pareto Genetic Algorithm (NPGA)
5.2.1 Applications of NPGA in Machine Learning
5.3 Pareto Archived Evolution Strategy (PAES)
5.3.1 Applications of PAES in Machine Learning
6 Conclusion
References
Machine Learning Assisted Metaheuristic Based Optimization of Mixed Suspension Mixed Product Removal Process
1 Introduction
2 Materials and Methods
2.1 MSMPR Crystallization Process
2.1.1 Mathematical Model
2.1.2 Optimization Setup
2.2 Support Vector Regression
3 Proposed Algorithm
3.1 Optimization of SVR Hyper-parameters (MOOP II)
4 Results and Discussions
4.1 Hyper-parameter Optimization
4.2 SVR Based Optimization of MSMPR
4.3 Comparison of Proposed Algorithm with Baseline Regression Techniques
4.4 Comparison of Proposed Algorithm with Fivefold Cross Validation
5 Conclusion
References
Machine Learning Based Intelligent RPL Attack Detection System for IoT Networks
1 Introduction
2 Related Works
3 Proposed Methodology
3.1 Description of Dataset
3.2 Intrusion Detection System
3.2.1 Eliminating Malicious Nodes from RPL
3.2.2 Feature Reduction
3.3 Proposed Neuro Genetic Algorithm for Classification
3.3.1 Reproduction and Mutation
3.3.2 Fitness Function
4 Results and Discussions
5 Conclusion
References
Shallow and Deep Evolutionary Neural Networks Applications in Solid Mechanics
1 Introduction
2 Optimal Design Using Material Mechanics
2.1 Laminates Usage
2.2 Decision Variable in Laminates Optimisation
2.3 Approaches to Fibre Composites Optimisation
2.4 Optimisation of Structures from Laminated Materials
3 Data-Driven Modelling Approaches
3.1 Data-Driven Evolutionary Algorithm
3.2 Algorithms for Evolutionary Metamodeling
3.2.1 Predator-Prey Genetic Algorithm
3.2.2 Reference Vector Evolutionary Algorithm (RVEA)
3.3 Data-Driven Model in Material Engineering
4 Optimisation of Mechanical Properties of Composite Beam Structures
4.1 Problem Description
4.2 Utilisation of Different Optimisation Approaches in the Simple Composite Beam Structures
4.3 Data-Driven Algorithm Applied to Curved Beams and Frames
4.3.1 Training Run
4.3.2 Optimisation Run
4.3.3 Results
5 Conclusions
References
Polymer and Nanocomposite Informatics: Recent Applications of Artificial Intelligence and Data Repositories
1 Introduction
2 Applications Areas
3 Databases and Tools for Polymer Informatics
4 Case Studies
5 Conclusion
References
Synergistic Combination of Machine Learning and Evolutionary and Heuristic Algorithms for Handling Imbalance in Biological and...
1 Introduction
2 Machine Learning Algorithms
2.1 Supervised Learning Algorithms
2.1.1 KNN (K Nearest Neighbors)
2.1.2 Naïve Bayes
2.1.3 SVM (Support Vector Machines)
2.1.4 Random Forest (RF)
2.2 Unsupervised Learning Algorithms
2.2.1 K-Means Clustering
2.3 Evaluation Measures for Classification Tasks
2.3.1 Area Under the Curve (AUC): Receiving Operating Characteristic (ROC)
2.3.2 Matthews Correlation Coefficient (MCC)
3 Imbalance of Data
3.1 Data Sampling Methods
3.1.1 Oversampling Techniques
3.1.2 Undersampling Techniques
3.1.3 Combination Techniques
4 Evolutionary and Heuristic Algorithms
4.1 Genetic Algorithm (GA)
4.2 Ant Colony Optimization
4.3 Particle Swarm Optimization (PSO)
4.4 Artificial Bee Colony (ABC)
4.5 Black Hole Algorithm
5 Case Studies Involving Evolutionary Algorithms Utilization for Imbalance Handling
5.1 General Algorithm Steps to Carry Out Undersampling
5.2 Identification of Cytokines Via an Improved GA [38]
5.3 DNA Microarray Imbalance Handling Via Ant Colony Optimization [30]
5.4 Protein Data Imbalance Handling Using Binary Particle Swarm Optimization [100]
5.5 Application of ABC for Some Imbalanced Bioinformatics Datasets [35]
5.6 Data Imbalance Handling for Proinflammatory Peptides and Diabetes Identification Using Binary Black Hole Algorithm
5.7 Data Imbalance Handling for Phase Separating Proteins Identification Using Combination of Generative Adversarial Networks ...
5.8 Some Imbalance Related Examples in Bioinformatics
6 Conclusion
References