Multi-disciplinary Trends in Artificial Intelligence : 15th International Conference, MIWAI 2022, Virtual Event, November 17–19, 2022, Proceedings

material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Preface Organization Contents Computing Nash Equilibrium of Crops in Real World Agriculture Domain 1 Introduction 2 Related Works 3 Non-cooperative Game 3.1 Strategic Form Game and Nash Equilibrium 3.2 Prisoner Dilemma 3.3 Cardinal vs Ordinal Utility 4 Complexity of the Problem 4.1 Typical Cases 4.2 Relation of Agent Payoffs 4.3 Case of 3 Agents and 2 Strategies 4.4 Case of 3 Agents and 3 Strategies 5 Searching for Nash Equilibrium 5.1 Control Loops 5.2 Algorithm for Examining Nash Equilibrium 5.3 Supporting Algorithms 6 Experiments and Results 6.1 Overview Result 6.2 Detailed Results 7 Conclusion References Evolutionary Feature Weighting Optimization and Majority Voting Ensemble Learning for Curriculum Recommendation in the Higher Education 1 Introduction 2 Material and Methods 2.1 Research Definition 2.2 Data Collection and Word Segmentation 2.3 Research Tools 3 Research Results 3.1 Model Performance Classified by Technique 3.2 Majority Voting Prototype Model 4 Research Discussion 5 Conclusion References Fuzzy Soft Relations-Based Rough Soft Sets Classified by Overlaps of Successor Classes with Measurement Issues 1 Introduction 2 Preliminaries 2.1 Some Basic Notions of Fuzzy Sets 2.2 Some Basic Notions of Soft Sets and Fuzzy Soft Relations 3 Main Results 3.1 Overlaps of Successor Classes via Fuzzy Soft Relations 3.2 Rough Soft Sets Based on Overlaps of Successor Classes 3.3 Measurement Issues 4 Conclusions References Helmet Detection System for Motorcycle Riders with Explainable Artificial Intelligence Using Convolutional Neural Network and Grad-CAM 1 Introduction 2 Related Works 2.1 Helmet Detection 2.2 Deep Learning and Convolution Neural Network 2.3 Histograms of Oriented Gradient (HOG) 2.4 Object Detection 2.5 Convolutional Neural Network 2.6 Explainable AI 2.7 Grad-CAM 3 Methodology 3.1 Data Collection and Preprocessing 3.2 Deep Convolution Neural Network 4 Experiment Setup and Results 4.1 Visualization and Explainable AI 5 Conclusion References Hierarchical Human Activity Recognition Based on Smartwatch Sensors Using Branch Convolutional Neural Networks 1 Introduction 2 Related Works 3 The Sensor-Based HAR Framework 3.1 WISDM-HARB Dataset 3.2 Data Pre-processing 3.3 Branch Convolutional Neural Network 3.4 Performance Measurement Criteria 4 Experiments and Results 4.1 Experiments 4.2 Experimental Results 5 Conclusions References Improving Predictive Model to Prevent Students’ Dropout in Higher Education Using Majority Voting and Data Mining Techniques 1 Introduction 2 Materials and Methods 2.1 Population and Sample 2.2 Data Acquisition Procedure 2.3 Model Construction Tools 2.4 Model Performance Evaluation Tools 3 Research Results 3.1 Generated Model Results 3.2 Majority Voting Prototype Model 4 Research Discussion 5 Conclusion 6 Limitation References LCIM: Mining Low Cost High Utility Itemsets 1 Introduction 2 Related Work 3 Problem Definition 4 The LCIM Algorithm 4.1 Search Space Exploration and Pruning Properties 4.2 The Cost-List Data Structure 4.3 The Algorithm 5 Experimental Evaluation 6 Conclusion References MaxFEM: Mining Maximal Frequent Episodes in Complex Event Sequences 1 Introduction 2 Problem Definition 3 The MaxFEM Algorithm 4 Experimental Evaluation 5 Conclusion References Method for Image-Based Preliminary Assessment of Car Park for the Disabled and the Elderly Using Convolutional Neural Networks and Transfer Learning 1 Introduction 2 Related Work 2.1 Manual Assessment of the Disabled Facilities 2.2 Computer Vision Techniques for Assessing Disabled Facilities or Accessibility 3 Research Methods 3.1 Data Collection and Labeling 3.2 Preliminary Assessment Method 3.3 Evaluating the Performance of the Method 4 Results and Discussion 5 Conclusion and Future Work References Multi-resolution CNN for Lower Limb Movement Recognition Based on Wearable Sensors 1 Introduction 2 Related Works 2.1 Types of Sensor Modalities 2.2 Deep Learning Approaches 3 The Sensor-Based HAR Framework 3.1 HARTH Dataset 3.2 Data Pre-processing 3.3 The Proposed Multi-resolution CNN Model 3.4 Performance Measurement Criteria 4 Experiments and Results 4.1 Experiments 4.2 Experimental Results 5 Conclusions References News Feed: A Multiagent-Based Push Notification System 1 Introduction 2 Review 3 Architecture, Internal and External Data Sources 4 Informative Multiagent-Based Personalized Data System 4.1 Components of Agents 4.2 Collective Agent 4.3 Analytic Agent 4.4 Dispense Agent 4.5 Farmer Agent 4.6 Scalable Design 4.7 Algorithm for Collecting Data 5 Results 5.1 Collecting Delay Time 5.2 Distributing Delay Time 5.3 Final Results 6 Conclusion References Optimizing the Social Force Model Using New Hybrid WOABAT-IFDO in Crowd Evacuation in Panic Situation 1 Introduction 2 Related Works 3 The Hybrid of WOABAT-IFDO and SFM Optimization Design Framework 3.1 Evacuation Time Validation 4 Result of Evacuation Time Hybrid WOABAT-IFDO in SFM vs Single SFM 4.1 Analysis of the Hypothesis for Evacuation Time Validation 5 Conclusions References Recognizing Driver Activities Using Deep Learning Approaches Based on Smartphone Sensors 1 Introduction 2 Related Works 3 Sensor-Based HAR Methodology 3.1 Driver Activity Dataset 3.2 Data Pre-processing 3.3 The Proposed DriveNeXt Architecture 4 Experiments and Research Findings 4.1 Research Setting 4.2 Research Findings 5 Conclusion and Future Works References Sentence-Level Sentiment Analysis for Student Feedback Relevant to Teaching Process Assessment 1 Introduction 2 Datasets 3 Preliminaries 3.1 Aspect-Based Keyword Corpus Development 3.2 Development of the Aspect Analyzer and the Sentiment Analyzer Using the Text Classification Technique 4 The Proposed Method 4.1 Pre-processing Student Comments and Text Representation 4.2 Identifying Aspect Class for Each Sentence Using the Aspect Analyzer 4.3 Assigning Sentence Polarity for Each Sentence Using the Sentiment Analyzer 4.4 Summarizing the Overall Sentiment Polarity of a Student Comment 5 Results 5.1 Evaluation of the Aspect Analyzer and the Sentiment Analyzer 5.2 Comparison of Proposed and Baseline Methods: In the Case of Modeling the Aspect Analyzer 6 Conclusion References Sentiment Analysis of Local Tourism in Thailand from YouTube Comments Using BiLSTM 1 Introduction 2 Related Work 2.1 Big Data Analytics 2.2 Deep Learning 2.3 Social Media Analytics (SMA) 2.4 Sentiment Analysis 3 Methodology 3.1 Data Collection 3.2 Sentiment Analysis 3.3 Summarizing Results 4 Evaluation 5 Conclusion References Stable Coalitions of Buyers in Real World Agriculture Domain 1 Introduction 2 Related Works 3 Real World Domain 3.1 Computing Values for a Group of Farmers 3.2 Computing Payoffs for Farmers 4 Stability in Coalition of Farmers 4.1 Coalition Formation 4.2 Kernel Solution Concept 5 Algorithms 5.1 Overview 5.2 Algorithm to Generate Coalitions 5.3 Algorithm to Verify Kernel 6 Experiments and Results 7 Conclusion References The Analysis of Explainable AI via Notion of Congruence 1 Introduction 2 Background 2.1 Abstract Argumentation 2.2 Assumption-Based Argumentation 2.3 Probabilistic Argumentation 3 Procedure 3.1 Translating BN Model to PABA Framework 3.2 Translating Argumentation Tree to PABA Framework 3.3 Establishing Notion of Congruence 4 Conclusion References Using Ensemble Machine Learning Methods to Forecast Particulate Matter (PM2.5) in Bangkok, Thailand 1 Introduction 2 Literature Review 3 Dataset Overview and Preparation 4 Research Methods 4.1 Seasonal ARIMA with Exogenous Covariates 4.2 Prophet Model 4.3 Regression Tree 4.4 Support Vector Regression 4.5 Artificial Neural Network 4.6 K-Nearest Neighbors (KNN) Regression 5 Results and Conclusions References Wearable Fall Detection Based on Motion Signals Using Hybrid Deep Residual Neural Network 1 Introduction 2 Related Works 2.1 Fall Detection System 2.2 Automatic Fall Detection by Using DL 3 Fall Detection Approach 3.1 FallAllD Dataset 3.2 Pre-processing of Data 3.3 Hybrid Deep Residual Neural Network 3.4 Interpretation Measurements 4 Experimental Results 5 Conclusion and Future Studies References Author Index This book constitutes the refereed proceedings of the 15th International Conference on Multi-disciplinary Trends in Artificial Intelligence, MIWAI 2022, held online on November 17-19, 2022. The 14 full papers and 5 short papers presented were carefully reviewed and selected from 42 submissions.