Advances In Pattern Recognition And Artificial Intelligence (Series On Language Processing, Pattern Recognition, And Intelligent Systems)

✍ Scribed by Marleah Blom (editor), Nicola Nobile (editor), Ching Yee Suen (editor)

• Publisher: WSPC
• Year: 2021
• Tongue: English
• Leaves: 277
• Category: Library

✦ Synopsis

This book includes reviewed papers by international scholars from the 2020 International Conference on Pattern Recognition and Artificial Intelligence (held online). The papers have been expanded to provide more details specifically for the book. It is geared to promote ongoing interest and understanding about pattern recognition and artificial intelligence. Like the previous book in the series, this book covers a range of topics and illustrates potential areas where pattern recognition and artificial intelligence can be applied. It highlights, for example, how pattern recognition and artificial intelligence can be used to classify, predict, detect and help promote further discoveries related to credit scores, criminal news, national elections, license plates, gender, personality characteristics, health, and more. Chapters include works centred on medical and financial applications as well as topics related to handwriting analysis and text processing, internet security, image analysis, database creation, neural networks and deep learning. While the book is geared to promote interest from the general public, it may also be of interest to graduate students and researchers in the field.

✦ Table of Contents

Contents
Preface
List of Contributors
Chapter 1: Enhanced Lesion Detection in Breast MRI Using Parallel and Cascaded Integration of Deep Learning Models
1 Introduction
2 Literature Review
3 Abnormality Detection Using Deep Learning
3.1 Faster R-CNN
3.2 YOLO
3.3 Grad-CAM
4 Integrated Approaches for Abnormality Detection
4.1 Cascaded Integration
4.2 Parallel Integration
5 Results and Discussions
6 Conclusion
References
Chapter 2: A Survey on Peripheral Blood Smear Analysis Using Deep Learning
1 Introduction
2 Related Works on Peripheral Blood Smear Analysis
2.1 Available Datasets
2.1.1 Synthetic Blood Smear Datasets
2.1.2 Non-synthetic Blood Smear and Blood Cell Datasets
2.2 Malaria Detection
2.3 Blood Cell Detection and Classification
2.4 Leukemia Diagnosis
3 Discussion and Future Directions
4 Conclusion
References
Chapter 3: End-to-End Generative Adversarial Network for Hand-Vein Recognition
1 Introduction
2 The Proposed Method
2.1 Labeling vein patterns
3 GAN Framework Structure
4 Experiment and Results
4.1 CASIA Dataset
4.2 PolyU Dataset
4.3 Verification Results
5 Conclusions
References
Chapter 4: Diabetic Retinopathy Analysis Based on Retinal Fundus Photographs via Deep Learning
1 Introduction
1.1 Harm of Diabetic Retinopathy
1.2 Methods for Diagnosing Diabetic Retinopathy
2 Acknowledgement of Diabetic Retinopathy
2.1 Pathogenesis of Diabetic Retinopathy
2.2 Retinal Images
2.3 Retinal Diseases
3 Diabetic Retinopathy Analysis Based on Deep Learning
3.1 Deep Learning
3.2 Our Methods for Fundus Disease Diagnosis
3.3 Balance of Training Dataset
3.4 Data Augmentation
4 Experiments
4.1 Dataset
4.2 Task
4.3 Data Augmentation
4.4 Accuracy and Loss
4.5 Analysis
5 Conclusion
References
Chapter 5: Pattern Recognition in Handwriting
Introduction
1 Factors that Affect Handwriting
1.1 External Factors
1.2 Internal Factors
2 What Happened to Handwriting Training?
3 Printing and Writing Patterns
3.1 Methods of Handwriting Instruction
3.2 Individualization
4 Master Patterns
5 Dimensions in Handwriting
6 Unified Patterns
6.1 Figure-Ground
7 Space–Form-Movement
7.1 Space
7.2 Form
7.3 Movement
7.4 Persona Writing Patterns
8 The Contraction and Release Continuum
9 A Brief History
10 In Summary
References
Chapter 6: A Deep Reinforcement Learning-based Study on Handwriting Difficulty Analysis
1 Introduction
2 Briefings on Bengali Conjuncts
3 Problem Formulation
4 Experimental Dataset Generation
5 Proposed Method
6 Experiments and Discussions
6.1 Database employed
6.2 Handwriting difficulty analysis
6.3 Comparison
7 Conclusion
7.1 Summary
7.2 Future work
7.3 Frontiers and future scope
Acknowledgments
References
Chapter 7: Gender Detection from Handwritten Documents
Using the Concept of Transfer Learning
1 Introduction
2 Related Works
3 Methodology
3.1 Pre-processing
3.2 Transfer-Learning
4 Experiments and Results
4.1 Dataset
4.2 Image classification results
5 Conclusions and Future Work
References
Chapter 8: The Wartegg Test
Introduction
1 The Test Itself
2 Scoring
3 Evolution of Scoring and Interpreting the Wartegg Test
4 Reliability and Validity of the Crisi Wartegg System
5 Potential Value and Applications
6 Can the Wartegg Test be Automated?
References
Chapter 9: A New Prediction Method for Credit Scoring Based on Sampling Reconstruction of Signal on Graph
1 Introduction
2 The Prediction Method Base on Sampling Reconstruction of Signal on Graph
2.1 Mathematical Formulation of the Prediction Problem
2.2 Solution of the Optimization Objection
2.3 A New Credit Prediction Algorithm
3 Experimental Results and Analysis
4 Conclusion
Achnowledgements
References
Chapter 10: Optimal Choices of Features in Image Analysis
1 Introduction
2 A feature: what for, whom for and how?
3 Diversity of features
3.1 Data structure choice: a mathematical point of view
3.2 Multisource
4 Feature properties
5 Selection versus Optimization
6 Conclusion
References
Chapter 11: A Comprehensive Unconstrained, License Plate Database
1 Introduction
1.1 Background Information
2 Database Overview
3 Data Collection
4 Ongoing Work
4.1 Labeling/Ground Truth
5 Conclusion
References
Chapter 12: Classification of Spanish Criminal News using Neural Networks
1 Introduction
2 Related work
3 Proposed approach
3.1 Word embeddings
3.2 Long Short-Term Memory
3.3 Regularisation
3.4 Max pooling
3.5 Attention mechanism
3.6 BERT-based model
4 Experimental results
4.1 Dataset description
4.2 Baseline
4.3 Bidirectional LSTM
4.4 Attention bidirectional LSTM
4.5 BERT
4.6 Results
5 Conclusions
Acknowledgments
References
Chapter 13: Predicting US Elections with Social Media and Neural Networks
1 Introduction
1.1 Motivation and Research Topic
1.1.1 Vocabulary of Characters
1.1.2 Histogram Representation
1.1.3 Data Encoding
1.1.4 Converting to Image Representation
1.1.5 Augmenting the Dataset
1.1.6 Shuffling Data
1.1.7 Splitting Data into Training, Validation, and Test Sets
1.1.8 Model Definition
1.1.9 GPU Acceleration
1.1.10 Randomized Data
1.1.11 Cross-Validation Results
1.1.12 T-Test of Predicted vs Actual Results
1.1.13 Election Results Predicted vs Actual
2 Conclusion
References
Chapter 14: Differences and Similarities learning for Unsupervised Feature Selection
1 Introduction
1.1 Preliminary
1.2 Related Works
2 The Proposed Algorithm FSDS
2.1 Exploiting Differences from Global Structure
2.2 Exploiting Similarities from Local Perspective
2.3 Sparseness Control
2.4 Subspace Feature Learning and Feature Combination
2.5 Iterative Algorithm of FSDS
2.6 Convergence Analysis
2.7 Discussion
3 The Proposed Algorithm FSDS2
3.1 FSDS2 Framework
3.2 Iterative Algorithm of FSDS2
3.3 Convergence Analysis
3.4 Discussion
4 Experiments
4.1 Datasets
4.2 Experiment Settings
4.3 Experimental Results
4.4 Convergence Study
4.5 Sensitivity Analysis
5 Conclusions
References
Index