دانلود کتاب Soft Computing and its Engineering Applications. 4th International Conference, icSoftComp 2022 Changa, Anand, India, December 9–10, 2022 Proceedings

فهرست مطالب :

Preface
Organization
Contents
Theory and Methods
NAARPreC: A Novel Approach for Adaptive Resource Prediction in Cloud
1 Introduction
1.1 Cloud Resource Prediction Challenges
1.2 Contribution
1.3 Organization
2 Related Work
3 Motivation
4 Background
4.1 Statistical Model: ARIMA
4.2 LSTM
5 System Model
6 Proposed Algorithm
6.1 Schematic Representation of NAARPreC
7 Experimental Results and Analysis
7.1 Dataset
7.2 Evaluation Metrics
7.3 Result Discussion
8 Conclusion and Future Scope
References
One True Pairing: Evaluating Effective Language Pairings for Fake News Detection Employing Zero-Shot Cross-Lingual Transfer
1 Introduction
2 Related Works
3 Methods
3.1 Definitions
3.2 Datasets
3.3 Cross-Lingual Language Models
3.4 Classification Algorithms
3.5 Pipeline Architecture
4 Results and Discussion
4.1 Monolingual Results
4.2 Cross-Lingual Results
5 Conclusion
References
FedCLUS: Federated Clustering from Distributed Homogeneous Data
1 Introduction
2 Related Work
3 System Model and Problem Definition
4 FedCLUS: The Proposed Horizontal Federated Clustering Method
4.1 Client Side Algorithm
4.2 Server Side Algorithm
5 Experiments and Results
5.1 Datasets and Experiments
5.2 Results of Performance Analysis
6 Conclusion
References
On Language Clustering: Non-parametric Statistical Approach
1 Introduction
2 Data Depth and Linguistic Clustering
3 Data Collection and Methodology
3.1 Data Collection
3.2 Distance Matrix
3.3 On Law of Large Numbers and Sampling Procedures
3.4 Interpreting the Distance Matrix
4 Application and Discussion
4.1 Outlier Detection
4.2 Unsupervised L1-Depth Based Clustering
4.3 Supervised Classifications and Other Possible Generalizations
5 Conclusion
References
Method Agnostic Model Class Reliance (MAMCR) Explanation of Multiple Machine Learning Models
1 Introduction
2 Related Works
3 Proposed Method
3.1 Models Building
3.2 Finding the Rashomon Set Models
3.3 Obtaining Model Reliance Values and Ranking Lists
3.4 Finding the Reference Explanation FileRef=\"538541_1_En_5_Figc_HTML.png\" Format=\"PNG\" Color=\"BlackWhite\" Type=\"Linedraw\" Rendition=\"HTML\" Height=\"29\" Resolution=\"300\" Width=\"46\"and Consistent Explanations
3.5 Computing the Weighted Grand Mean (θ)
3.6 Method Agnostic Model Class Reliance (MAMCR) Explanation
4 Experiments and Results
4.1 Discussion
5 Conclusion
References
Association Rules Based Feature Extraction for Deep Learning Classification
1 Introduction
2 Literature Review
2.1 Association Rules for Feature Selection
2.2 Association Rules for Result Analysis
2.3 Other Methods
3 Background
3.1 Association Rule Mining
3.2 Neural Networks
3.3 Deep Learning Residual Networks, ResNet
4 AR with ResNet Based Classifier
4.1 Association Rules
4.2 Classification Models
4.3 Datasets
5 Experiments and Results
5.1 AR1
5.2 AR2
5.3 Classifier Implementation and Results
5.4 Results of Breast Cancer and Dermatology Datasets
6 Conclusion
References
Global Thresholding Technique for Basal Ganglia Segmentation from Positron Emission Tomography Images
1 Introduction
2 Related Works
3 Proposed Approach
3.1 Preparing Dataset
3.2 Ground Truth Construction
3.3 Thresholding Segmentation
3.4 Performance Evaluation
4 Experiments
4.1 Results and Discussion
4.2 Limits and Future Work
5 Conclusion
References
Oversampling Methods to Handle the Class Imbalance Problem: A Review
1 Introduction
2 Oversampling Methods Used
3 Dataset
4 Experimental Results
5 Conclusion
References
Analog Implementation of Neural Network
1 Introduction
2 Literature Survey
3 Neuron Network Architecture
3.1 Multiplier and Adder
3.2 Approach
4 Implementation
4.1 31 Positive Weight Matrix Multiplication
4.2 3x1 Negative Weight Matrix Multiplication
4.3 33 Weight Matrix Multiplication
4.4 2 Layer of 33 Weight Matrix Multiplication
5 Performance Comparison
6 Conclusion
References
Designing Fog Device Network for Digitization of University Campus
1 Introduction
1.1 Network Architecture of Fog-integrated Cloud
2 The Proposed Work
2.1 Set of Network Entities
2.2 Constants
2.3 Functions
2.4 Decision Variables
2.5 The Mathematical Model
2.6 The Weighted Sum Multi-objective Optimization Method
3 The Experimental Evaluation
3.1 Experimental Input
3.2 Experimental Results
3.3 Analysis of Results
4 Conclusion
References
Word Sense Disambiguation from English to Indic Language: Approaches and Opportunities
1 Introduction
2 Resources Required for Disambiguation
2.1 Machine Readable Dictionaries/Thesaurus
2.2 WordNet
2.3 Corpus
3 Variants of Word Sense Disambiguation Work
3.1 Lexical Sample (or Target Word or One Word) WSD
3.2 All Word WSD
4 Related Work
5 Proposed Approach for WSD
5.1 Architecture of the Proposed WSD Model
5.2 Implementation Details
6 Result Discussion
7 Conclusion and Future Directions
References
Explainable AI for Predictive Analytics on Employee Attrition
1 Introduction
2 Related Works
3 Materials and Methods
3.1 Dataset Description
3.2 Data Preprocessing
3.3 Explainable AI
3.4 SHAPley Additive Explanations
4 Results and Discussion
5 Conclusions
References
Graph Convolutional Neural Networks for Nuclei Segmentation from Histopathology Images
1 Introduction
2 Literature Review
3 Dataset Description
4 Research Methodology
4.1 Data Augmentation
4.2 Image Normalization
4.3 GraphSegNet Architecture
5 Experimental Setup
5.1 Implementation Details
5.2 Evaluation Metrics
6 Results
6.1 Model Performance
7 Conclusion
References
Performance Analysis of Cache Memory in CPU
1 Introduction
2 Literature Review
3 Methodology for Benchmark Program
3.1 CPU Performance
4 Computer Specifications
5 Benchmark Logic
6 Results and Graph Analysis
7 Conclusion and Future Scope
References
Extraction of Single Chemical Structure from Handwritten Complex Chemical Structure with Junction Point Based Segmentation
1 Introduction
2 Critical Review
3 Segmentation Procedure
3.1 Pre-processing
3.2 Skeletonization
3.3 Junction Point Detection
3.4 Segmentation
4 Experiment
4.1 Input
4.2 Process
4.3 Result and Discussion
5 Conclusion
References
Automatic Mapping of Deciduous and Evergreen Forest by Using Machine Learning and Satellite Imagery
1 Introduction
2 Study Area and Data
3 Methodology
3.1 Random Forest
3.2 k-Nearest Neighbor
4 Results and Discussion
5 Conclusions
References
Systems and Applications
RIN: Towards a Semantic Rigorous Interpretable Artificial Immune System for Intrusion Detection
1 Introduction
2 Explainable Artificial Intelligent Driven Intrusion Detections
3 RIN– Rigorous XAI Driven Intrusion Detection
3.1 Rigorous XAI
3.2 M&M – Discretize Continuous Features
3.3 Architecture of RIN
4 Evaluation Results of RIN
4.1 Feature Discretization
4.2 Semantic Rigorous Explanations
5 Conclusion and Future Research Challenges
References
SVRCI: An Approach for Semantically Driven Video Recommendation Incorporating Collective Intelligence
1 Introduction
2 Related Work
3 Proposed Work
4 Implementation and Performance Evaluation
5 Conclusion
References
Deep Learning Based Model for Fundus Retinal Image Classification
1 Introduction
2 Used Methods
2.1 Basic of Convolutional Neural Network (CNN)
2.2 Working of Machine Learning Based Binary Classifiers
3 Experimental Setup
3.1 Dataset
3.2 Details of Proposed CNN Model
3.3 Parameters for Training of CNN Model
3.4 Experimental Platform
4 Experimental Results and Discussion
5 Conclusion
References
Reliable Network-Packet Binary Classification
1 Introduction
2 Related Work
2.1 Port-number Based Techniques
2.2 Deep Packet Inspection(DPI) or Payload Based Techniques
2.3 Machine Learning-Based Techniques
3 Proposed Approach
3.1 Dimensionality Reduction Techniques
3.2 Dataset
3.3 Pre-processing and Labelling
3.4 Architecture
4 Results and Analysis
4.1 Comparison
5 Conclusion and Future Work
References
SemKnowNews: A Semantically Inclined Knowledge Driven Approach for Multi-source Aggregation and Recommendation of News with a Focus on Personalization
1 Introduction
2 Related Works
3 Proposed Architecture
4 Implementation
5 Performance Evaluation and Results
6 Conclusion
References
Extraction and Analysis of Speech Emotion Features Using Hybrid Punjabi Audio Dataset
1 Introduction
2 Related Work
3 SER Process
3.1 Audio Emotional Dataset
3.2 Extraction of Features
3.3 Selection of Features
3.4 Recognition of Emotions
4 The Performance Evaluation
4.1 Neural Network Configuration
4.2 Experimental Results
4.3 Comparison with other SER Systems
5 Conclusion and Future Scope
References
Human Activity Recognition in Videos Using Deep Learning
1 Introduction
2 Literature Survey
3 Methodology
3.1 Dataset
3.2 Preprocessing
3.3 Model Evaluation and Training
4 Results and Analysis
5 Conclusion and Future Work
References
Brain Tumor Classification Using VGG-16 and MobileNetV2 Deep Learning Techniques on Magnetic Resonance Images (MRI)
1 Introduction
2 Methodology
2.1 Visual Geometry Group (VGG)
2.2 MobileNet
2.3 Evaluation Parameters for Deep Learning Models
3 Results and Analysis
4 Conclusions
References
Five-Year Life Expectancy Prediction of Prostate Cancer Patients Using Machine Learning Algorithms
1 Introduction
2 Literature Review
3 Methodology
3.1 Data Collection
3.2 Experimental Setup
3.3 Data Preparation
3.4 Correlation Analysis
3.5 Data Splitting
3.6 Machine Learning Model Building
3.7 Hyperparameter Tuning
3.8 Performance Evaluation
3.9 Cross-Validation:
4 Results and Discussion
5 Conclusion
References
An Ensemble MultiLabel Classifier for Intra-Cranial Haemorrhage Detection from Large, Heterogeneous and Imbalanced Database
1 Introduction
2 Dataset Description
3 Proposed Ensemble MultiLabel Classifier for ICH Detection
3.1 Data Pre-processing
3.2 SX-DNN: The Proposed Model for Automatic ICH Detection
4 Experimental Results
4.1 Training
4.2 Evaluation
4.3 Results
5 Discussion
6 Conclusion
References
A Method for Workflow Segmentation and Action Prediction from Video Data - AR Content
1 Introduction
2 Related Work
3 Solution Approach
3.1 Workflow Creation from Video Content Data
3.2 Training Sequential Models to Identify Actions and Descriptions
3.3 Real Time Action Detection and Suggestions
4 Application in Manufacturing Domain
5 Conclusion and Next Steps
References
Convolutional Neural Network Approach for Iris Segmentation
1 Introduction
2 Related Work
3 Proposed System
4 Experiment and Results
5 Conclusion and Future Work
References
Reinforcement Learning Algorithms for Effective Resource Management in Cloud Computing
1 Introduction
2 Literature Survey
3 Experiment
3.1 Experiment Configuration and Simulated Environment
3.2 Experiment Dataset
3.3 VM Configuration
3.4 RL Rewards and Q-Table
4 Results and Implications
4.1 Results and Implications with Respect to Resource Scheduling
4.2 Results and Implications with Respect to Fault Tolerance
5 Conclusions
References
Corpus Building for Hate Speech Detection of Gujarati Language
1 Introduction
2 Related Forum and Dataset
3 Dataset and Collection
4 Methodology
4.1 Data Preprocessing
4.2 Data Annotation
5 Experiments
6 Result and Discussion
7 Conclusion
References
Intrinsic Use of Genetic Optimizer in CNN Towards Efficient Image Classification
1 Introduction
2 Related Work
3 Proposed Work
3.1 Input
3.2 Convolution
3.3 Genetic Optimizer for CNN (GOCNN)
4 Experimental Analysis
4.1 Dataset Description
5 Discussion
6 Conclusion
References
A Software System for Smart Course Planning
1 Introduction
2 Early Systems
3 Course Level Characteristics
3.1 Characteristic 1
3.2 Characteristic 2
4 The SCPS Software
4.1 Test Case 1
4.2 Test Case 2
5 Conclusion
References
Meetei Mayek Natural Scene Character Recognition Using CNN
1 Introduction
2 Related Work
3 Meetei Mayek Natural Scene Character Extraction and Database Creation
3.1 Maximally Stable Extremal Regions (MSER) Detection
3.2 Geometric Filtering
3.3 Filtering According to Stroke Width
3.4 Filtering Considering Distance
3.5 Meetei Mayek Natural Scene Character Database Creation
4 Proposed CNN for Meetei Mayek Natural Scene Character Classification
4.1 CNN Architecture
5 Experimental Results
6 Conclusion and Future Works
References
Utilization of Data Mining Classification Technique to Predict the Food Security Status of Wheat
1 Introduction
2 Current Situation of Wheat Production and Consumption
3 Related Works
4 Research Objectives
5 The Proposed Model and Framework
5.1 The Proposed Model to Predict FSSW
5.2 The Proposed Framework to Predict and Manage FSSW
6 Case Study
6.1 Phase of Food Balance of Wheat Dataset (FBWD)
6.2 Prediction Process Phase
6.3 Phase of the Research Results and Model Accuracy
7 Results and Discussion
7.1 Comparative Study Between the MPFSSW and the Previous Works
7.2 Recommendations
8 Conclusions and Future Work
References
Hybrid Techniques
QoS-Aware Service Placement for Fog Integrated Cloud Using Modified Neuro-Fuzzy Approach
1 Introduction
1.1 Motivation
2 Related Work
3 Proposed Model
3.1 Architecture of ANFIS
3.2 Learning Method of Anfis
4 Result and Discussion
5 Conclusion
References
Optimizing Public Grievance Detection Accuracy Through Hyperparameter Tuning of Random Forest and Hybrid Model
1 Introduction
2 Preliminaries
2.1 Hyperparameter Tuning on Random Forest
2.2 Hybrid Approach in Machine Learning
2.3 Signum Function
3 Data Collection and Pre-processing
4 Previous Experiments
4.1 Phase 1
4.2 Phase 2
5 Experiment Phase 3
5.1 Random Forest Hyperparameter Tuning
5.2 Hybrid Algorithm for Class Determination
6 Results
7 Conclusion
8 Future Scope
References
Author Index