Modern Approaches in Machine Learning and Cognitive Science: A Walkthrough: Latest Trends in AI, Volume 2 (Studies in Computational Intelligence, 956)

✍ Scribed by Vinit Kumar Gunjan; Jacek M. Zurada

• Tongue: English
• Leaves: 503
• Category: Library

✦ Table of Contents

Preface
Contents
Using CNN to Predict Regressive STD Drug Efficacy Score
1 Introduction
2 Use of CNN for Sentiment Analysis
3 Advantages and Limitations
4 Methodology
4.1 Preprocessing
4.2 Word Embeddings
4.3 Convolutional Neural Network
5 Architecture
6 Dataset
7 Result
8 Conclusion
References
Emotion Recognition in Hindi Speech Using CNN-LSTM Model
1 Introduction
2 Background
2.1 Emotion Recognition in Speech
2.2 Neural Networks
2.3 Applications of Neural Networks
2.4 MFCC
3 Dataset
4 Feature Extraction and Preprocessing
5 Proposed Methodology
5.1 CNN Model
5.2 ANN Model
5.3 CNN—LSTM Model
6 Results and Conclusion
References
Refinery Profit Planning via Evolutionary Many-Objective Optimization
1 Introduction
2 Overview of NSGA-III
2.1 Measures of Convergence and Diversity
3 Problem Formulation
4 Results and Discussion
5 Conclusions
References
A Deep Learning Technique for Image Inpainting with GANs
1 Introduction
2 Literature Survey
3 Methodology
3.1 Training
3.2 Testing
4 Algorithm
4.1 Flow of Control
5 Implementation
6 Results and Discussion
7 Conclusion
References
A Comparative Study on Distributed File Systems
1 Introduction
2 History of File Systems
3 Literature Survey
4 Google File System
5 Hadoop File System
6 Other File Systems and Services
6.1 OpenAFS
6.2 OpenIO
6.3 Google Cloud Storage
6.4 MapR File System (MapR FS)
7 Comparison
8 Conclusion
References
An Organized Approach for Analysis of Diabetic Nephropathy Images Using Watershed and Contrast Adaptive Thresholding
1 Introduction
2 Need and Importance of Research Problem
3 Destinations and Scope
4 Plan of Work and Methodology
5 Algorithm
6 Results and Analysis
6.1 Watershed Method
7 Conclusion
References
A Literature Survey on Identification of Asthma Using Different Classifier and Clustering Techniques
1 Introduction
2 Study on Preprocessing and Have Selection in Asthma Disease
2.1 Pre-processing Strategy
2.2 Feature Selection
2.3 Highlight Choice Methods
3 Survey on Different Classifier of Asthma Using Neural Networks
4 Brief Review of Clustering Techniques
4.1 K-Nearest Neighbor Classifier
4.2 Bolster Vector Machine (SVM)
4.3 Choice Tree Classifier
5 Conclusion
References
Adaptation and Evolution of Decision Support Systems—A Typological Survey
1 Introduction
1.1 Background
1.2 Main Focus of the Article
2 Evolution of DSS Framework
2.1 Communications-Driven
2.2 Data-Driven
2.3 Model-Driven
2.4 Knowledge-Driven
2.5 Document-Driven
3 Evolution of DSS Applications
3.1 Business Enterprise Management
3.2 Healthcare
3.3 Agriculture and Natural Resources Management
4 Evolution of DSS Architecture
4.1 Interactive Decision Support
4.2 Web-Based Decision Support
4.3 Artificial Intelligence (AI) Based Decision Support
5 The Gap in Literature
6 Conclusion
References
Real-Time Implementation of Brain Emotional Controller for Sensorless Induction Motor Drive with Adaptive System
1 Introduction
2 Architecture and Computational Model of  Brain Emotional Controller
3 MRAS Speed Estimator  for IM drive
4 SVM-DTC for Induction Motor
5 Simulation Results
6 Experimental Results
7 Conclusions
Appendix 1
References
Student Performance Prediction—A Data Science Approach
1 Introduction
2 Literature Review
3 Machine Learning Models for Experimentation
3.1 Methodology-I: Analysis and Prediction of First Year Performance
3.2 Methodology-II: Predicting Semester/Course Wise Low Performing Students
4 Results Analysis
5 Conclusions and Future Work
References
HARfog: An Ensemble Deep Learning Model for Activity Recognition Leveraging IoT and Fog Architectures
1 Introduction
2 Related Work
3 Background Technologies
4 System Architecture for HAR
4.1 Sensors at the Edge
4.2 Gateway Layer
4.3 Cloud Layer
4.4 Deep Learning Module
5 Implementation and Evaluation
6 Conclusion
References
Performance Evaluation and Identification of Optimal Classifier for Credit Card Fraudulent Detection
1 Introduction
2 Data Analysis Methodology and Classifier Selection
3 Data Preparation and Performance Evaluation of Classification Algorithms
3.1 Logistic Regression
3.2 Linear Discriminant Analysis (LDA)
3.3 K-Nearest Neighbors (K-NN)
3.4 Classification and Regression Trees (CART)
3.5 Support Vector Machine (SVM)
3.6 eXtreme Gradient Boosting (XGBoost)
3.7 Random Forest (RF)
4 Result and Discussion
5 Conclusion
References
Potential Use-Cases of Natural Language Processing for a Logistics Organization
1 Introduction
2 Motivation Behind NLP in Logistics
2.1 NLP in Breaking Language Barrier
2.2 NLP in Contract Management
2.3 NLP in Order Management
2.4 NLP in Processing of Logistics Textual Data
2.5 NLP for Sentiment Analysis and Customer Satisfaction
2.6 NLP in Operational Procurement
2.7 NLP in Information Extraction
2.8 NLP in Improving Efficiency
2.9 NLP in Transportation Management
2.10 NLP in Automation
3 Possible Challenges of NLP
4 Logistics Tasks
4.1 Operational Procurement
4.2 SCM Parties Communication
4.3 Logistics Documentation
4.4 Customer Satisfaction and Service
4.5 Predictive Risk Management
4.6 Contract Management and Drafting
4.7 Supply Chain Planning
4.8 Supplier Relationship Management (SRM)
4.9 Auditing and Compliance Check
5 Some of the Major Approaches of NLP for Logistics Tasks
5.1 Rule-Based NLP Approach
5.2 Machine Translation
5.3 Text Similarity
5.4 Word Embedding
5.5 POS Tagging
5.6 Keyword Extraction
5.7 Deep Learning
5.8 Chatbots
5.9 Reinforcement Learning
5.10 Text Classification
5.11 Relationship Extraction
5.12 Language Modelling
5.13 Document Summarization
5.14 Word Sense Disambiguation (WSD)
5.15 Ontology
5.16 Sentimental Analysis
5.17 Question Answering
5.18 Topic Modelling
6 Finding and Conclusions
References
Partial Consensus and Incremental Learning Based Intrusion Detection System
1 Introduction
1.1 Organization of Paper
2 Related Work
3 Proposed Solution
3.1 Algorithm
3.2 Learning
3.3 Selection of Nodes
3.4 Updating the Weights
3.5 Consensus
3.6 Parameters
4 Performance Analysis
5 Conclusion
References
AI Enabled Context Sensitive Information Retrieval System
1 Introduction
1.1 Experiment Dataset
1.2 Related Work
2 Technical Approach
3 Results
4 Conclusion and Future Work
Appendix
References
Personalization of News for a Logistics Organisation by Finding Relevancy Using NLP
1 Introduction
2 Motivation Behind NLP for News Relevancy
3 Literature Review
4 System Overview
5 Materials and Methods
5.1 Data Collection
5.2 Model Building
5.3 Front End
6 Experiment and Results
7 Challenges Encountered
8 Conclusion
References
AI Model Compression for Edge Devices Using Optimization Techniques
1 Introduction
2 Background Study
2.1 Convolutional Neural Network (CNN)
2.2 CIFAR-10
2.3 Deep Learning
2.4 ResNet-50
2.5 Pruning
2.6 Quantization
3 Related Work
4 Proposed Work
4.1 Working of the Proposed System
4.2 Removal of Unimportant Weights in Each Layer of the Neural Network
4.3 Conversion of Float32 to Int8
5 Results
5.1 Accuracy Drop versus Sparsity
5.2 Distribution of Weights in CNNs Layer
5.3 Comparison Between Baseline Model and Proposed Model
6 Conclusion
References
An Empirical Study and Analysis of Various Electroencephalography (EEG) Artefact Removal Methods
1 Introduction
2 Characteristics of the EEG
3 Denoising Techniques for EEG Signals
3.1 Independent Component Analysis (ICA)
3.2 Principal Component Analysis (PCA)
3.3 Regression Method
3.4 Common Spatial Patterns (CSP)
3.5 Singular Spectrum Analysis (SSA)
3.6 Common Average Referencing (CAR)
3.7 Surface Laplacian (SL)
3.8 Adaptive Filtering
4 Relative Analysis
5 Conclusion
References
Chatbot via Machine Learning and Deep Learning Hybrid
1 Introduction
2 Literature Review
3 Proposed Methodology
3.1 Data Preparation
3.2 Hybrid Multi-level Model (HMLM)
4 Dataset Description
5 Results and Discussion
6 Conclusions
References
Prediction Intervals for Macroeconomic Variables Using LSTM Based LUBE Method
1 Introduction
2 Literature Survey
3 Proposed Methodology
3.1 Modified LUBE Method with LSTM
4 Experimental Design
4.1 Data Sets Description
4.2 Evaluation Metrics
5 Results and Discussion
5.1 CPI Food and Beverages
5.2 CPI Headline
5.3 CPI Fuel and Light
6 Conclusion
References
Intelligent Character Recognition with Shared Features from Resnet
1 Introduction
2 Background
2.1 Resnet
2.2 Feature Pyramidal Networks
2.3 Levenshtein Distance
3 Proposed Methodology
4 Dataset Description
4.1 IAM Forms Dataset
4.2 Synthesized GMB (SGMB) Dataset
4.3 Student Answer Sheets Dataset
4.4 Train—Test Split and Dictionary Creation
4.5 Evaluation Metrics
5 Results and Discussions
6 Conclusion
References
SVM and Naïve Bayes Models for Estimation of Key Process Variables in Nuclear Power Plant
1 Introduction
2 Literature Review
3 Relevant Data Mining Algorithms
3.1 Naïve Bayesian Classifier Model
3.2 Support Vector Machines
4 The Process System
5 Model Building
6 Model Evaluation
7 Conclusion
References
Face Recognition Using Transfer Learning on Facenet: Application to Banking Operations
1 Introduction
2 Background
2.1 Histogram of Oriented Gradients (HOG)
2.2 Transfer Learning
2.3 Triplet Loss and Facenet
3 Proposed Methodology
4 Dataset Description
4.1 Evaluation Metrics
5 Results and Discussions
6 Conclusions
References
Deep Learning Chatbot with Feedback
1 Introduction
2 Literature Review
3 Proposed Methodology
3.1 Data Preparation
3.2 Model Overview
4 Dataset Description
5 Results and Discussion
6 Conclusions
References
A Non-monotonic Activation Function for Neural Networks Validated on Benchmark Tasks
1 Introduction
2 Non-monotone Activation Function
3 Tasks and Experiment Design
3.1 Benchmark Tasks
4 Experiment Design
5 Results and Discussions
6 Conclusions
References
Analysis of Approaches for Automated Glaucoma Detection and Prediction System
1 Introduction
1.1 Glaucoma
1.2 Glaucoma Detection System
1.3 Glaucoma Prediction System
2 Related Work
2.1 Supervised Learning
2.2 Unsupervised Learning
2.3 Reinforcement Learning
2.4 Deep Learning
3 Biological Survey
4 Research Gaps and Findings
5 Solutions to the Research Gaps
6 Conclusion
References
Experiences in Machine Learning Models for Aircraft Fuel Flow and Drag Polar Predictions
1 Introduction
2 Data Visualization and Principal Component Analysis
2.1 Principal Component Analysis
3 Fuel Flow Predictions
3.1 Prediction Model
3.2 Model Evaluation
3.3 Drag Polar Computation
3.4 Applications of Fuel Consumption NN Models
4 Conclusions
References
Wildlife Video Captioning Based on ResNet and LSTM
1 Introduction
2 Related Work
3 Proposed Model
3.1 Convolutional Neural Networks
3.2 Recurrent Neural Networks
3.3 Hierarchical Attentive Encoder
3.4 Globally and Locally Aligned Cross-Modal Attentive Decoder
4 Experimental Setup
4.1 Dataset Used
4.2 Preprocessing
4.3 Evaluation Metrics
4.4 Training Details
5 Results and Discussion
6 Conclusions
References
Enhancement of Degraded Images via Fuzy Intensification Model
1 Introduction
2 Literature Survey
3 Methodology
4 Experimental Investigations
5 Conclusion
References
Design and Investigation of the Performance of Composite Spiral Antenna for Direction Finding Applications
1 Introduction
2 Concept of Frequency Independent Antennas
2.1 Basic Principles of Direction Finding Systems
3 Composite Spiral Antennas
4 Results and Discussion
4.1 Simulated Radiation Patterns
4.2 Comparison on the Investigated Results
4.3 Gain of the Antenna
5 Conclusions
References
Learning Based Approach for Subtle Maintenance in Large Institutions
1 Introduction
2 Case Study
3 Objective
4 Related Work
5 Overview of Approach
6 Working
6.1 Pre-processing of Data
6.2 Classification of Complaints Using Artificial Neural Network
6.3 Prediction of Future Number of Complaints Using Recurrent Neural Network
7 Experiment and Results
8 Conclusion
References
Temporal Localization of Topics Within Videos
1 Introduction
2 Related Work
3 Proposed System
3.1 Methodology
4 Implementation and Results
5 Conclusion
6 Future Enhancement
References
Fast Training of Deep Networks with One-Class CNNs
1 Introduction
2 Proposed Approach
3 Experimentation
3.1 4Face Database
3.2 Caltech-101 Image Database
3.3 17Flowers Database
4 Conclusion and Future Work
References
Recognition of Isolated English Words of E-Lecture Video Using Convolutional Neural Network
1 Introduction
2 Related Work
3 Proposed Work
4 Implementation
4.1 Feature Extraction
4.2 Convolutional Neural Network
5 Results
6 Conclusion and Future Work
References
Indoor Object Location Finding Using UWB Technology
1 Introduction
2 UWB Positioning
2.1 Comparison with Another Positioning Technology
3 Theory
3.1 Ranging
4 Implementation
4.1 System Overview
4.2 Hardware
4.3 Software
4.4 Server/GUI
5 Results and Discussion
6 Conclusion
References
Design and Tuning of Improved Current Predictive Control for PMSM
1 Introduction
2 Concept of Predictive Current Control
3 ANFIS Controller
4 Design of Hybrid Controller:
5 Simulation Results
6 Conclusion
References
Parametric Analysis of Texture Classification Using Modified Weighted Probabilistic Neural Network (MWPNN)
1 Introduction
2 Probabilistic Neural Network
2.1 Input Layer
2.2 Hidden Layer/Pattern Layer
2.3 Summation Layer
2.4 Output Layer
3 Weighted Probabilistic Neural Network
3.1 Supervised Labelling Mechanism
3.2 WPNN Algorithm
3.3 Sensitivity Analysis
3.4 Bench Mark Classifiers Used in the Comparison
3.5 Multi-Layer Perceptron (MLP)
3.6 Navie Bayes Classifier (NB)
4 Results and Discussions
5 Conclusion
References
Modeling and Simulation of Automatic Centralized Micro Grid Controller
1 Introduction
2 Structure Modeling
2.1 Modeling of the Systems Sources
3 Modeling of Converters
3.1 Boost Converter Modeling
3.2 DFIG Controller Modeling
3.3 Modeling of Bi-directional Converter
4 Modeling of ACMC
4.1 Algorithm for Load Shedding
4.2 Algorithm for Plug and Play
5 Case Study
6 FUZZY Logic Controller
7 Simulation Results
8 Conclusion
References
The VLSI Realization of Sign-Magnitude Decimal Multiplication Efficiency
1 Introduction
2 Literature Review
3 Current System
4 Proposed System
4.1 Recoding of Multiplier’s Digits
4.2 Partial Product Generation
4.3 Partial Product Reduction (PPR)
5 Implementation of the Proposed Multiplier in Fir Filter
5.1 Fir Filtre
6 Result
7 Conclusion
References
Image Encryption Algorithms Using Machine Learning and Deep Learning Techniques—A Survey
1 Introduction
2 Machine Learning Algorithms
2.1 Image Encryption Using Chaotic Based Artificial Neural Network
2.2 Novel Priority Based Document Image Encryption with Mixed Chaotic Systems Using Machine Learning Approach
2.3 Learnable Image Encryption
2.4 Encrypted Image Retrieval System: A Machine Learning Approach
2.5 Machine Learning Classification Over Encrypted Data
2.6 Reversible Data Hiding Scheme During Encryption Using Machine Learning
3 Deep Learning Algorithms
3.1 An Image Compression and Encryption Scheme Based on Deep Learning
3.2 Batch Image Encryption Using Generated Deep Features Based on Stacked Auto Encoder Network
3.3 Research on Iris Image Encryption Based on Deep Learning
3.4 Privacy-Preserving Distributed Deep Learning via Homomorphic Re-Encryption
3.5 DLEDNet: A Deep Learning-Based Image Encryption and Decryption Network for Internet of Medical Things
4 Conclusion
References