Intelligent Systems and Applications : Proceedings of the 2023 Intelligent Systems Conference (IntelliSys) Volume 2 / edited by Kohei Arai.

Format:

Book

Author/Creator:

Arai, Kohei.

Series:

Lecture Notes in Networks and Systems, 2367-3389 ; 823

Language:

English

Subjects (All):

Computational intelligence.

Automatic control.

Robotics.

Automation.

Artificial intelligence.

Computational Intelligence.

Control, Robotics, Automation.

Artificial Intelligence.

Local Subjects:

Computational Intelligence.

Control, Robotics, Automation.

Artificial Intelligence.

Physical Description:

1 online resource (890 pages)

Edition:

1st ed. 2024.

Place of Publication:

Cham : Springer Nature Switzerland : Imprint: Springer, 2024.

Summary:

The book is a unique collection of studies involving intelligent systems and applications of artificial intelligence in the real world to provide solutions to most vexing problems. IntelliSys received an overwhelming 605 papers which were put under strict double-blind peer-review for their novelty, originality and exhaustive research. Finally, 227 papers were sieved and chosen to be published in the proceedings. This book is a valuable collection of all the latest research in the field of artificial intelligence and smart systems. It provides a ready-made resource to all the readers keen on gaining information regarding the latest trends in intelligent systems. It also renders a sneak peek into the future world governed by artificial intelligence.

Contents:

Intro

Preface

Contents

YOLO-Based Object Detection in Industry 4.0 Fischertechnik Model Environment

1 Introduction

2 The Fischertechnik Learning Factory

3 Dataset Preparation

4 YOLO Models

4.1 YOLOv1

4.2 YOLOv3

4.3 YOLOv5

5 Detection Results

6 Summary and Conclusions

References

Towards End-to-End Escape in Urban Autonomous Driving Using Reinforcement Learning

2 Materilas and Methods

2.1 Related Work

2.2 Experiments Setup

3 Results

3.1 Settings Evaluation

4 Conclusions

MS-ETL: An Architecture for the Multiple Data Source Extraction, Transformation, and Load Applied to Solar Flares Data

2 Concepts and Background

3 Multiple Data Source Extraction, Transformation, and Load (MS-ETL) Architecture

3.1 The Extraction of the Series of Solar Satellite Images

3.2 Extraction of Solar Flares Data

4 Results and Discussions

5 Conclusion and Future Works

Wind Turbine Data-Driven Intelligent Fault Detection

2 Intelligent Control Methodology

3 Proposed Strategy

4 Results and Discussion

5 Conclusion

Enhancing Decision Support Systems for the Energy Sector with Sustainable Artificial Intelligence Solutions

2 Background

3 I-NERGY Framework

3.1 Overview

3.2 I-NERGY Use Cases

3.3 I-NERGY Assets and AIoD Experiments

4 Conclusion and Next Steps

Pre-trained Deep Learning Models for Chest X-Rays' Classification: Views and Age-Groups

2 Pre-trained Models' Architecture

2.1 VGG16

2.2 ResNet50

2.3 Inception V3

2.4 Xception

2.5 MobileNetV2

2.6 DenseNet201

2.7 EfficientNetB7

3 Methodology

3.1 Data Preparation

3.2 Implementation

3.3 Evaluation

4 Results

5 Discussion.

6 Conclusion

Impact of Gender and Chest X-Ray View Imbalance in Pneumonia Classification Using Deep Learning

2 Architecture and Methodology

2.1 Architecture

2.2 Datasets

3.1 Training

3.2 Testing

3.3 Gender-Based Testing

3.4 View-Based Testing

4 Discussion

5 Conclusion and Future Work

Gesture Recognition of Filipino Sign Language Using Convolutional and Long-Short Term Memory Neural Network

2 Background and Related Works

2.1 Sign Language Recognition (SLR) Systems

2.2 Challenges on FSL Research

3.1 Proposed Design System Architecture

3.2 Data Collection and Preparation

3.3 Training

Index Tracking Via Learning to Predict Market Sensitivities

2 Related Work

2.1 Index Tracking via Machine Learning

2.2 Estimation of Market Sensitivities

3 Preliminaries

3.1 Notations

3.2 Single-Factor Model

3.3 Deep-Learning Models

4 Proposed Methods

4.1 Overview of Our Methods

4.2 Prediction of Market Sensitivities via Deep Learning Models

4.3 Portfolio Construction

5 Experiments

5.1 Input Data Structure

5.2 Temporal Data Splitting To Preclude Look-Ahead Bias

5.3 Experimental Settings

5.4 Evaluation of the Historical Estimation

5.5 Prediction of Market Sensitivities and Alphas

5.6 Performance of the Portfolios

6 Conclusion

Deep Learning Models for Inventory Decisions: A Comparative Analysis

2 Literature Review

2.1 Deep Learning for Demand Forecasting

2.2 Methods for Achieving Single-Period Inventory Decisions

3 Deep Learning Methods

3.1 Multi-layer Perceptron (MLP)

3.2 Convolutional Neural Networks (CNN)

3.3 Recurrent Neural Networks (RNN).

3.4 Long Short-Term Memory (LSTM)

4 Single-Period Inventory Optimization: The Newsvendor Problem

5 Case Study

5.1 Dataset Overview and Preparation

5.2 Model Building and Implementation

5.3 Newsvendor Problem Parameters

5.4 Results and Discussion

Detecting Standard Library Functions in Obfuscated Code

3 Methods

3.1 Data Augmentation Through Obfuscation

3.2 Data Preprocessing

3.3 PV-DM Voting Classifier

3.4 Graph Metadata Classifier

5 Results

6 Conclusions and Future Work

How Starting Points and Representations Affect Software Modularisation: An Empirical Analysis

1.1 Software Modularisation

1.2 Optimisation

1.3 Solution Representation

1.4 Search Space and Contributions

2 Research Questions

3 Software Clustering

3.1 Our Approach to Optimising Software

3.2 Initial Clustering Arrangements

4 Datasets

5 Experimental Procedure

5.1 Run Time

6 Results

6.1 What is the Best Representation?

6.2 What is the Best Starting Point?

6.3 What is the Best Combination of Representation and Starting Point?

6.4 Does Size Matter?

7 Threat to Validity

8 Conclusions and Future Work

ASR Bundestag: A Large-Scale Political Debate Dataset in German

2 German Datasets for Speech Recognition

3 Data Acquisition and Processing

3.1 Acquirement of Data

3.3 Alignment Methods

4 The German Bundestag Dataset

4.1 Alignment Based on Timestamps and CTC

4.2 Alignment Based on Silence Intervals and Speaker Diarization

5 Speech Recognition Results

6 Discussion

7 Related Work

8 Conclusion and Outlook

References.

Prototype App Mobile for Real Time American Sign Language Recognition Based on Deep Learning

2 Methodology

3.1 Data Capture

3.2 Data Pre-processing

3.3 Convolutional Neuronal Network

3.4 Classification and Training

3.5 Evaluation

3.6 App Mobile Prototype

5 Recommendations

A Novel DNN-Based IDS System Combined with an LR-GA Method to Detect Attacks

3 The Proposed Model

3.1 Data Encoding

3.2 Data Clustering

3.3 Data Feature Selections

3.4 Deep Neural Network-Based IDS Model

4 Validations and Discussions

4.1 KDD 99 and IoT-23 Datasets Overview

4.2 Exprimental Results

Identification of Potato Virus Y in Potato Plants Using Deep Learning and GradCAM Verification

2 State of the Art

3 Materials and Methods

3.1 Data Collection

3.2 Data Processing

3.3 Classification Models

3.4 Hyperparameters

3.5 Evaluation Metrics

3.6 Feature Extraction Capabilities

4 Experiments and Results

4.1 Experiment 1: Ablation Study

4.2 Experiment 2: Feature Extraction Capabilities

5 Discussion, Conclusion and Future Work

5.1 Discussion

5.2 Conclusion

5.3 Future Work

Using Simulated Data for Deep-Learning Based Real-World Apple Detection

3.1 Simulating Apple Orchards in Unity

3.2 Generating Data in Unity

3.3 Real-World Datasets

3.4 Apple Detection Using YOLOv5

4.1 Experiment 1: Performance of Augmentations

4.2 Experiment 2: Performance of Simulated Data Versus a Traditional Dataset

4.3 Experiment 3: Performance of a Hybrid Dataset

5 Discussion, Conclusion and Future Work.

5.1 Discussion

Attention-Based Recurrent Neural Network for Multicriteria Recommendations

2.1 Multicriteria Recommender Systems

2.2 Sequence-Aware Recommender Systems

3.1 Recurrent Neural Networks

3.2 Long-Short Term Memory

4 Proposed Method

4.1 LSTM-RNN Model

4.2 Attention Mechanism

5 Experimental Evaluation

5.1 Data Set

5.2 Attention

5.3 Experimental Results and Discussion

Improved Technique for Dimensionality Reduction: Star and Quasar Classification with Typical Testors

2 Testor Theory

2.1 YYC Algorithm

3.1 Dataset

3.2 Basic Matrix

3.3 Typical Testor Calculation

3.4 Validation

4.1 Basic Matrix and Typical Testors

4.2 SVM Model Training with All the Features

4.3 SVM Model Training with Typical Testor Reduction

4.4 SVM Model Training with PCA Reduction

4.5 Training Precision

4.6 Analysis

5 Conclusions

5.1 Future Work

AgriScanNet-18: A Robust Multilayer CNN for Identification of Potato Plant Diseases

3.2 Convolutional Neural Networks

3.3 Proposed AgriScanNet-18 Multilayer Convolutional Neural Network

3.4 Experimental Setup

4 Results and Analysis

4.1 Evaluation Measures

4.2 Web Application Results

Automatic Optimization-Based Methods in Machine Learning: A Systematic Review

2 Literature Review Methodology

3 Hyperparameters Optimization

4 CNN's Structural Optimization

5 Dataset-Based Optimization

7 Conclusion

Interpolation and Prediction of Piezometric Multivariate Time Series Based on Data Augmentation and Transformers.

Notes:

Description based on publisher supplied metadata and other sources.

ISBN:

9783031477249

3031477243

OCLC:

1431193562