Note: Intro -- Foreword -- References -- Preface -- Contents -- 1 Introduction to Fusion of Machine Learning Paradigms -- 1.1 Editorial -- References -- Part I Recent Application Areas of Fusion of Machine Learning Paradigms -- 2 Artificial Intelligence as Dual-Use Technology -- 2.1 Introduction -- 2.2 What Is DUT -- 2.3 AI: Concepts, Models and Technology -- 2.4 Agent-Based AI and Autonomous System -- 2.4.1 Basic Model of Agent-Based AI -- 2.4.2 Conceptual Model of Autonomous Weapon System -- 2.5 Dual-Use Technology and DARPA -- 2.5.1 Historical View and Role of DARPA -- 2.5.2 DARPA's Contribution to DUT R& -- D on AI -- 2.6 DARPA-Like Organizations in Major Countries -- 2.7 Dual-Use Dilemma -- 2.8 Concluding Remarks -- References -- 3 Diabetic Retinopathy Detection Using Transfer and Reinforcement Learning with Effective Image Preprocessing and Data Augmentation Techniques -- 3.1 Introduction -- 3.2 Background -- 3.2.1 Deep Learning for Diabetic Retinopathy -- 3.2.2 Image Preprocessing Techniques -- 3.2.3 Reinforcement Learning and Deep Learning -- 3.3 Data Augmentation Techniques -- 3.3.1 Traditional Data Augmentation -- 3.3.2 SMOTE-Based Data Augmentation -- 3.3.3 Data Augmentation Using Generative Adversarial Networks -- 3.4 Datasets of Eye Fundus Images -- 3.5 Transfer Learning Experiments -- 3.5.1 Dataset -- 3.5.2 Image Preprocessing -- 3.5.3 Image Augmentation -- 3.5.4 Deep Learning Experiments -- 3.5.5 Reinforcement Learning Experiments -- 3.6 Conclusion and Future Work -- References -- 4 A Novel Approach for Non-linear Deep Fuzzy Rule-Based Model and Its Applications in Biomedical Analyses -- 4.1 Introduction -- 4.2 Method -- 4.2.1 Preliminaries -- 4.2.2 Hierarchical Fuzzy Structure -- 4.2.3 Stacked Deep Fuzzy Rule-Based System (SD-FRBS) -- 4.2.4 Adaptation of the First-Order TSK Structure in SD-FRBS. , 4.2.5 Concatenated Deep Fuzzy Rule-Based System (CD-FRBS) -- 4.3 Data Description and Results -- 4.3.1 MIMIC-III Dataset -- 4.3.2 SD-FRBS as a Multivariate Regressor for Granger Causality Estimation-In EEG Connectivity Index Extraction -- 4.3.3 CD-FRBS in Staging Depression Severity -- 4.4 Discussion and Conclusion -- 4.4.1 Suggested Future Works -- References -- 5 Harmony Search-Based Approaches for Fine-Tuning Deep Belief Networks -- 5.1 Introduction -- 5.2 Theoretical Background -- 5.2.1 Deep Belief Networks -- 5.2.2 Harmony Search -- 5.3 Methodology -- 5.3.1 Datasets -- 5.3.2 Experimental Setup -- 5.4 Experimental Results -- 5.5 Conclusions -- References -- 6 Toward Smart Energy Systems: The Case of Relevance Vector Regression Models in Hourly Solar Power Forecasting -- 6.1 Introduction -- 6.2 Relevance Vector Regression -- 6.3 RVR Based Day Ahead Forecasting -- 6.4 Results -- 6.5 Conclusion -- References -- 7 Domain-Integrated Machine Learning for IC Image Analysis -- 7.1 Introduction -- 7.2 Hierarchical Multi-classifier System -- 7.2.1 Architecture of Hierarchical Multi-classifier System -- 7.2.2 Result and Discussion on Case Study -- 7.3 Deep Learning with Pseudo Labels -- 7.3.1 Methodology -- 7.3.2 Application to IC Image Analysis -- 7.4 Conclusions and Future Works -- References -- Part II Applications that Can Clearly Benefit from Fusion of Machine Learning Paradigms -- 8 Fleshing Out Learning Analytics and Educational Data Mining with Data and ML Pipelines -- 8.1 Introduction -- 8.2 Data and ML Pipelines -- 8.3 Related Work -- 8.4 An Automated EDM and LA Methodology -- 8.4.1 A Data Pipeline Scenario -- 8.4.2 An ML Pipeline Scenario -- 8.5 Experiments and Results -- 8.6 Conclusions and Future Work -- References -- 9 Neural Networks Based Throughput Estimation of Short Production Lines Without Intermediate Buffers -- 9.1 Introduction. , 9.2 Data Sets of i-Stage Production Line Problems -- 9.3 Deep Learning and Multilayer Perceptron -- 9.4 Experimental Process of Deep Learning Approach -- 9.5 Results of Deep Learning Approach -- 9.6 Conclusions -- References.

Note: Intro -- Foreword -- Preface -- Contents -- 1 Applications of Learning and Analytics in Intelligent Systems -- References -- Learning and Analytics in Intelligent Medical Systems -- 2 A Comparison of Machine Learning Techniques to Predict the Risk of Heart Failure -- 2.1 Introduction -- 2.2 Theoretical Background -- 2.3 Data and Methods -- 2.3.1 Dataset -- 2.3.2 Evaluation Process -- 2.3.3 Weka -- 2.4 Overview of Proposed Systems -- 2.4.1 Logistic Regression -- 2.4.2 Decision Tree -- 2.4.3 Random Forest -- 2.4.4 K-Nearest Neighbor -- 2.4.5 Artificial Neuronal Network -- 2.4.6 SVM -- 2.4.7 Naïve Bayes -- 2.4.8 OneR -- 2.4.9 ZeroR -- 2.4.10 Hybrid -- 2.5 Comparison Results -- 2.6 Conclusions -- References -- 3 Differential Gene Expression Analysis of RNA-seq Data Using Machine Learning for Cancer Research -- 3.1 Introduction -- 3.2 Materials and Methods -- 3.2.1 RNAseq -- 3.2.2 Classical Approach -- 3.2.3 Machine Learning -- 3.2.4 Comparative Workflow -- 3.3 Code and Results of an Analysis with Real Data -- 3.3.1 Loading Packages -- 3.3.2 Loading and Searching the Data from TCGA -- 3.3.3 Patient Selection -- 3.3.4 Dependent Variable Definition -- 3.3.5 Biological Gene Filter -- 3.3.6 Graphics -- 3.3.7 Classical Statistical Analysis -- 3.3.8 Machine Learning Analysis -- 3.4 Conclusions -- References -- 4 Machine Learning Approaches for Pap-Smear Diagnosis: An Overview -- 4.1 Introduction -- 4.2 Cervical Cancer and Pap-Test -- 4.3 The Pap-Smear Databases -- 4.3.1 A Basic Data Analysis of New Data -- 4.4 The Used Methodologies -- 4.4.1 Adaptive Network-Based Fuzzy Inference System (ANFIS) -- 4.4.2 Artificial Neural Networks -- 4.4.3 Heuristic Classification -- 4.4.4 Minimum Distance Classification -- 4.4.5 Hard C-Means Clustering -- 4.4.6 Fuzzy C-Means Clustering -- 4.4.7 Gustafson and Kessel Clustering -- 4.4.8 k-Nearest Neighborhood Classification. , 4.4.9 Weighted k-Nearest Neighborhood Classification -- 4.4.10 Tabu Search -- 4.4.11 Genetic Programming -- 4.4.12 Ant Colony -- 4.5 The Pap-Smear Classification Problem -- 4.5.1 Classification with ANFIS -- 4.5.2 Heuristic Classification Based on GP -- 4.5.3 Classification Using Defuzzification Methods -- 4.5.4 Direct and Hierarchical Classification -- 4.5.5 Classification Using Feed-Forward Neural Network -- 4.5.6 Nearest Neighborhood Classification Based on GP Feature Selection -- 4.5.7 Nearest Neighborhood Classification Using Tabu Search for Feature Selection -- 4.5.8 Nearest Neighborhood Classification Using ACO for Feature Selection -- 4.5.9 Minimum Distance Classifier -- 4.6 Conclusion and Future Work -- References -- Learning and Analytics in Intelligent Power Systems -- 5 Multi-kernel Analysis Paradigm Implementing the Learning from Loads Approach for Smart Power Systems -- 5.1 Introduction -- 5.2 Background -- 5.2.1 Kernel Machines -- 5.2.2 Gaussian Processes -- 5.3 Multi-kernel Paradigm for Load Analysis -- 5.3.1 Problem Statement -- 5.3.2 Multi-kernel Paradigm -- 5.4 Results -- 5.4.1 Problem Statement -- 5.4.2 Further Results -- 5.5 Conclusion and Future Work -- References -- 6 Conceptualizing and Measuring Energy Security: Geopolitical Dimensions, Data Availability, Quantitative and Qualitative Methods -- 6.1 Preamble -- 6.1.1 Structure of Chapter -- 6.2 Review of Energy Security Literature -- 6.2.1 Brief History of Energy Security -- 6.2.2 Defining Security and Energy Security -- 6.2.3 Energy Security Since the 20th Century -- 6.2.4 Energy Security and Geopolitics -- 6.2.5 Dimensions of Energy Security -- 6.3 Methodology -- 6.3.1 Research Questions -- 6.4 Analyses and Results -- 6.4.1 Milestone Time Periods -- 6.4.2 Data -- 6.4.3 Measuring Energy Security -- 6.4.4 Creating a Geopolitical Energy Security Index. , 6.4.5 Using Cluster Analysis -- 6.4.6 Looking at Case Studies of Key Countries -- 6.4.7 Carrying Out Interviews of Energy Experts -- 6.4.8 Forecasting Energy Security -- 6.5 Closing Comments -- References -- Learning and Analytics in Performance Assessment -- 7 Automated Stock Price Motion Prediction Using Technical Analysis Datasets and Machine Learning -- 7.1 Introduction -- 7.2 Technical Analysis Synopsis -- 7.3 Machine Learning Component -- 7.3.1 SVM Algorithm -- 7.3.2 Adaptive Boost Algorithm -- 7.4 System Implementation -- 7.4.1 System Structure -- 7.4.2 Training Data Set -- 7.4.3 Selection of Machine Learning Algorithm and Implementation -- 7.4.4 Android Client Application -- 7.5 System Evaluation -- 7.6 Conclusions and Future Work -- References -- 8 Airport Data Analysis Using Common Statistical Methods and Knowledge-Based Techniques -- 8.1 Introduction -- 8.2 Literature Review -- 8.3 Airport Data Analysis -- 8.3.1 Data Collection and Cleansing -- 8.3.2 Case Study Description and Scope of Current Analysis -- 8.3.3 Demand Seasonality -- 8.3.4 International Passenger Connectivity Matrix -- 8.3.5 Weekly and Daily Airport Operating Patterns -- 8.3.6 Airplane Types and Associated Runway Length Requirements -- 8.4 Conclusions -- References -- 9 A Taxonomy and Review of the Network Data Envelopment Analysis Literature -- 9.1 Introduction -- 9.2 DMU's Internal Network Structures and Assessment Paradigms -- 9.3 Assessment Paradigms -- 9.3.1 Independent Assessments -- 9.3.2 Joint Assessments -- 9.4 Classification of Network DEA Studies -- 9.5 Conclusion -- References -- Learning and Analytics in Intelligent Safety and Emergency Response Systems -- 10 Applying Advanced Data Analytics and Machine Learning to Enhance the Safety Control of Dams -- 10.1 Introduction -- 10.2 The Data Lifecycle in the Safety Control of Concrete Dams. , 10.2.1 Raw Data Collection -- 10.2.2 Processing and Data Storage -- 10.2.3 Data Quality Assessment and Outlier Detection -- 10.2.4 Data Analysis and Dam Safety Assessment Based on Quantitative Interpretation Models -- 10.2.5 Data Analysis and Dam Safety Assessment Based on Machine Learning Models -- 10.3 Data Analysis and Data Prediction Using Deep Learning Models-An Overview -- 10.4 Adopted Problem Solving Process-The Design Science Research Methodology -- 10.5 Proposed Methodology-Adding Value to the Interpretation of the Monitored Dam Behaviour Through the Use of Deep Learning Models -- 10.6 Demonstration and Evaluation-Assessment and Interpretation of the Monitored Structural Behaviour of a Concrete Dam During Its Operation Phase -- 10.6.1 The Case Study-The Alto Lindoso Dam -- 10.6.2 The Dataset-Horizontal Displacements Measured by the Pendulum Method -- 10.6.3 Main Results and Discussion -- 10.7 Final Remarks -- References -- 11 Analytics and Evolving Landscape of Machine Learning for Emergency Response -- 11.1 Introduction -- 11.1.1 Emergency Management -- 11.1.2 Machine Learning -- 11.1.3 Scope and Organizations -- 11.2 Applications of Machine Learning in Emergency Response -- 11.2.1 Machine Learning Techniques for Emergency Management Cycles -- 11.2.2 Event Prediction -- 11.2.3 Warning Systems -- 11.2.4 Event Detection and Tracking -- 11.2.5 Situational Awareness -- 11.2.6 Emergency Evaluation -- 11.2.7 Crowdsourcing -- 11.3 Analysis of Emergency Data -- 11.3.1 Big Data in Emergency Management -- 11.3.2 Data Collection -- 11.3.3 Information Extraction and Filtering -- 11.3.4 Data Integration -- 11.3.5 Applications for Data Analysis in Emergency -- 11.4 Challenges and Opportunities of Machine Learning in Response -- 11.4.1 Data Collection -- 11.4.2 Information Extraction -- 11.4.3 Data Filtering -- 11.4.4 Data Integration. , 11.5 Crowdsourcing in Emergency Management -- 11.5.1 Crowdsourcing with Machine Learning for Emergency Management -- 11.5.2 Example: Crowdsourcing and Machine Learning for Tracking Emergency -- 11.6 Conclusions -- References -- Learning and Analytics in Intelligent Social Media -- 12 Social Media Analytics, Types and Methodology -- 12.1 Social Networks and Analytics -- 12.1.1 Descriptive Analytics -- 12.1.2 Diagnostic Analytics -- 12.1.3 Predictive Analytics -- 12.1.4 Prescriptive Analytics -- 12.2 Introduction to Social Network Mining -- 12.3 Data Structure -- 12.3.1 Structured Data -- 12.3.2 Semi-structured Data -- 12.3.3 Unstructured Data -- 12.4 Data Quality -- 12.4.1 Noise -- 12.4.2 Outliers -- 12.4.3 Missing Values -- 12.4.4 Duplicate Data -- 12.5 Data Preprocessing -- 12.5.1 Aggregation -- 12.5.2 Discretization -- 12.5.3 Feature Selection -- 12.5.4 Feature Extraction -- 12.5.5 Sampling -- 12.6 Network Modeling -- 12.6.1 Real World Networks -- 12.6.2 Random Graphs -- 12.6.3 Small World Model -- 12.6.4 Preferential Attachment Model -- 12.7 Network Schemas -- 12.7.1 Multi-relational Network with Single Typed Objects -- 12.7.2 Bipartite Network -- 12.7.3 Star-Schema Network -- 12.7.4 Multiple-hub Network -- 12.8 Task Categorization -- 12.9 Machine Learning -- 12.9.1 Supervised Learning -- 12.9.2 Unsupervised Learning -- 12.10 Conclusions -- References -- 13 Machine Learning Methods for Opinion Mining In text: The Past and the Future -- 13.1 Introduction -- 13.2 Terminology -- 13.3 Early Projects -- 13.4 The Fascinating Opportunities that Sentiment Analysis Raises -- 13.5 Natural Language Processing for Sentiment Analysis -- 13.5.1 Affective Information for Sentiment Analysis -- 13.5.2 Corpora Annotated for Sentiment Analysis Tasks -- 13.5.3 Distributional Semantics and Sentiment Analysis. , 13.6 Traditional Models Based on Lexica and Feature Engineering.

Note: Intro -- Foreword -- Further Reading -- Preface -- Contents -- 1 Introduction to Advances in Machine Learning/Deep Learning-Based Technologies -- 1.1 Editorial Note -- 1.2 Book Summary and Future Volumes -- References -- Part I Machine Learning/Deep Learning in Socializing and Entertainment -- 2 Semi-supervised Feature Selection Method for Fuzzy Clustering of Emotional States from Social Streams Messages -- 2.1 Introduction -- 2.2 The FS-EFCM Algorithm -- 2.2.1 EFCM Execution: Main Steps -- 2.2.2 Initial Parameter Setting -- 2.3 Experimental Results -- 2.3.1 Dataset -- 2.3.2 Feature Selection -- 2.3.3 FS-EFCM at Work -- 2.4 Conclusion -- References -- 3 AI in (and for) Games -- 3.1 Introduction -- 3.2 Game Content and Databases -- 3.3 Intelligent Game Content Generation and Selection -- 3.3.1 Generating Content for a Language Education Game -- 3.4 Conclusions -- References -- Part II Machine Learning/Deep Learning in Education -- 4 Computer-Human Mutual Training in a Virtual Laboratory Environment -- 4.1 Introduction -- 4.1.1 Purpose and Development of the Virtual Lab -- 4.1.2 Different Playing Modes -- 4.1.3 Evaluation -- 4.2 Background and Related Work -- 4.3 Architecture of the Virtual Laboratory -- 4.3.1 Conceptual Design -- 4.3.2 State-Transition Diagrams -- 4.3.3 High Level Design -- 4.3.4 State Machine -- 4.3.5 Individual Scores -- 4.3.6 Quantization -- 4.3.7 Normalization -- 4.3.8 Composite Evaluation -- 4.3.9 Success Rate -- 4.3.10 Weighted Average -- 4.3.11 Artificial Neural Network -- 4.3.12 Penalty Points -- 4.3.13 Aggregate Score -- 4.4 Machine Learning Algorithms -- 4.4.1 Genetic Algorithm for the Weighted Average -- 4.4.2 Training the Artificial Neural Network with Back-Propagation -- 4.5 Implementation -- 4.5.1 Instruction Mode -- 4.5.2 Evaluation Mode -- 4.5.3 Computer Training Mode -- 4.5.4 Training Data Collection Sub-mode. , 4.5.5 Machine Learning Sub-mode -- 4.6 Training-Testing Process and Results -- 4.6.1 Training Data -- 4.6.2 Training and Testing on Various Data Set Groups -- 4.6.3 Genetic Algorithm Results -- 4.6.4 Artificial Neural Network Training Results -- 4.7 Conclusions -- References -- 5 Exploiting Semi-supervised Learning in the Education Field: A Critical Survey -- 5.1 Introduction -- 5.2 Semi-supervised Learning -- 5.3 Literature Review -- 5.3.1 Performance Prediction -- 5.3.2 Dropout Prediction -- 5.3.3 Grade Level Prediction -- 5.3.4 Grade Point Value Prediction -- 5.3.5 Other Studies -- 5.3.6 Discussion -- 5.4 The Potential of SSL in the Education Field -- 5.5 Conclusions -- References -- Part III Machine Learning/Deep Learning in Security -- 6 Survey of Machine Learning Approaches in Radiation Data Analytics Pertained to Nuclear Security -- 6.1 Introduction -- 6.2 Machine Learning Methodologies in Nuclear Security -- 6.2.1 Nuclear Signature Identification -- 6.2.2 Background Radiation Estimation -- 6.2.3 Radiation Sensor Placement -- 6.2.4 Source Localization -- 6.2.5 Anomaly Detection -- 6.3 Conclusion -- References -- 7 AI for Cybersecurity: ML-Based Techniques for Intrusion Detection Systems -- 7.1 Introduction -- 7.1.1 Why Does AI Pose Great Importance for Cybersecurity? -- 7.1.2 Contribution -- 7.2 ML-Based Models for Cybersecurity -- 7.2.1 K-Means -- 7.2.2 Autoencoder (AE) -- 7.2.3 Generative Adversarial Network (GAN) -- 7.2.4 Self Organizing Map -- 7.2.5 K-Nearest Neighbors (k-NN) -- 7.2.6 Bayesian Network -- 7.2.7 Decision Tree -- 7.2.8 Fuzzy Logic (Fuzzy Set Theory) -- 7.2.9 Multilayer Perceptron (MLP) -- 7.2.10 Support Vector Machine (SVM) -- 7.2.11 Ensemble Methods -- 7.2.12 Evolutionary Algorithms -- 7.2.13 Convolutional Neural Networks (CNN) -- 7.2.14 Recurrent Neural Network (RNN) -- 7.2.15 Long Short Term Memory (LSTM). , 7.2.16 Restricted Boltzmann Machine (RBM) -- 7.2.17 Deep Belief Network (DBN) -- 7.2.18 Reinforcement Learning (RL) -- 7.3 Open Topics and Potential Directions -- 7.3.1 Novel Feature Representations -- 7.3.2 Unsupervised Learning Based Detection Systems -- References -- Part IV Machine Learning/Deep Learning in Time Series Forecasting -- 8 A Comparison of Contemporary Methods on Univariate Time Series Forecasting -- 8.1 Introduction -- 8.2 Related Work -- 8.3 Theoretical Background -- 8.3.1 ARIMA -- 8.3.2 Prophet -- 8.3.3 The Holt-Winters Seasonal Models -- 8.3.4 N-BEATS: Neural Basis Expansion Analysis -- 8.3.5 DeepAR -- 8.3.6 Trigonometric BATS -- 8.4 Experiments and Results -- 8.4.1 Datasets -- 8.4.2 Algorithms -- 8.4.3 Evaluation -- 8.4.4 Results -- 8.5 Conclusions -- References -- 9 Application of Deep Learning in Recurrence Plots for Multivariate Nonlinear Time Series Forecasting -- 9.1 Introduction -- 9.2 Related Work -- 9.2.1 Background on Recurrence Plots -- 9.2.2 Time Series Imaging and Convolutional Neural Networks -- 9.3 Time Series Nonlinearity -- 9.4 Time Series Imaging -- 9.4.1 Dimensionality Reduction -- 9.4.2 Optimal Parameters -- 9.5 Convolutional Neural Networks -- 9.6 Model Pipeline and Architecture -- 9.6.1 Architecture -- 9.7 Experimental Setup -- 9.8 Results -- 9.9 Conclusion -- References -- Part V Machine Learning in Video Coding and Information Extraction -- 10 A Formal and Statistical AI Tool for Complex Human Activity Recognition -- 10.1 Introduction -- 10.2 The Hybrid Framework-Formal Languages -- 10.3 Formal Tool and Statistical Pipeline Architecture -- 10.4 DATA Pipeline -- 10.5 Tools for Implementation -- 10.6 Experimentation with Datasets to Identify the Ideal Model -- 10.6.1 KINISIS-Single Human Activity Recognition Modeling -- 10.6.2 DRASIS-Change of Human Activity Recognition Modeling -- 10.7 Conclusions. , References -- 11 A CU Depth Prediction Model Based on Pre-trained Convolutional Neural Network for HEVC Intra Encoding Complexity Reduction -- 11.1 Introduction -- 11.2 H.265 High Efficiency Video Coding -- 11.2.1 Coding Tree Unit Partition -- 11.2.2 Rate Distortion Optimization -- 11.2.3 CU Partition and Image Texture Features -- 11.3 Proposed Methodology -- 11.3.1 The Hierarchical Classifier -- 11.3.2 The Methodology of Transfer Learning -- 11.3.3 Structure of Convolutional Neural Network -- 11.3.4 Dataset Construction -- 11.4 Experiments and Results -- 11.5 Conclusion -- References.

Note: Intro -- Foreword -- Preface -- Contents -- 1 Machine Learning Paradigms: Introduction to Deep Learning-Based Technological Applications -- 1.1 Editorial Note -- References -- Part IDeep Learning in Sensing -- 2 Vision to Language: Methods, Metrics and Datasets -- 2.1 Introduction -- 2.2 Challenges in Image Captioning -- 2.2.1 Understanding and Predicting `Importance' in Images -- 2.2.2 Visual Correctness of Words -- 2.2.3 Automatic Evaluation Metrics -- 2.2.4 Image Specificity -- 2.2.5 Natural-Sounding Descriptions -- 2.3 Image Captioning Models and Their Taxonomy -- 2.3.1 Example Lookup-Based Models -- 2.3.2 Generation-based Models -- 2.4 Assessment of Image Captioning Models -- 2.4.1 Human Evaluation -- 2.4.2 Automatic Evaluation Metrics -- 2.4.3 Distraction Task(s) Based Methods -- 2.5 Datasets for Image Captioning -- 2.5.1 Generic Captioning Datasets -- 2.5.2 Stylised Captioning Datasets -- 2.5.3 Domain Specific Captioning Datasets -- 2.6 Applications of Visual Captioning -- 2.6.1 Medical Image Captioning -- 2.6.2 Life-Logging -- 2.6.3 Commentary for Sports' Videos -- 2.6.4 Captioning for Newspapers -- 2.6.5 Captioning for Assistive Technology -- 2.6.6 Other Applications -- 2.7 Extensions of Image Captioning to Other Vision-to-Language Tasks -- 2.7.1 Visual Question Answering -- 2.7.2 Visual Storytelling -- 2.7.3 Video Captioning -- 2.7.4 Visual Dialogue -- 2.7.5 Visual Grounding -- 2.8 Conclusion and Future Works -- References -- 3 Deep Learning Techniques for Geospatial Data Analysis -- 3.1 Introduction -- 3.2 Deep Learning: A Brief Overview -- 3.2.1 Deep Learning Architectures -- 3.2.2 Deep Neural Networks -- 3.2.3 Convolutional Neural Network (CNN) -- 3.2.4 Recurrent Neural Networks (RNN) -- 3.2.5 Auto-Encoders (AE) -- 3.3 Geospatial Analysis: A Data Science Perspective -- 3.3.1 Enabling Technologies for Geospatial Data Collection. , 3.3.2 Geospatial Data Models -- 3.3.3 Geospatial Data Management -- 3.4 Deep Learning for Remotely Sensed Data Analytics -- 3.4.1 Data Pre-processing -- 3.4.2 Feature Engineering -- 3.4.3 Geospatial Object Detection -- 3.4.4 Classification Tasks in Geospatial Analysis -- 3.5 Deep Learning for GPS Data Analytics -- 3.6 Deep Learning for RFID Data Analytics -- 3.7 Conclusion -- References -- 4 Deep Learning Approaches in Food Recognition -- 4.1 Introduction -- 4.2 Background -- 4.2.1 Popular Deep Learning Frameworks -- 4.3 Deep Learning Methods for Food Recognition -- 4.3.1 Food Image Datasets -- 4.3.2 Approach #1: New Architecture Development -- 4.3.3 Approach #2: Transfer Learning and Fine-Tuning -- 4.3.4 Approach #3: Deep Learning Platforms -- 4.4 Comparative Study -- 4.4.1 New Architecture Against Pre-trained Models -- 4.4.2 Deep Learning Platforms Against Each Other -- 4.5 Conclusions -- References -- Part IIDeep Learning in Social Media and IOT -- 5 Deep Learning for Twitter Sentiment Analysis: The Effect of Pre-trained Word Embedding -- 5.1 Introduction -- 5.2 Related Work -- 5.3 Evaluation Procedure -- 5.3.1 Datasets -- 5.3.2 Data Preprocessing -- 5.3.3 Pre-trained Word Embeddings -- 5.3.4 Deep Learning -- 5.4 Comparative Analysis and Discussion -- 5.5 Conclusion and Future Work -- References -- 6 A Good Defense Is a Strong DNN: Defending the IoT with Deep Neural Networks -- 6.1 Introduction -- 6.2 State of the Art in IoT Cyber Security -- 6.3 A Cause for Concern: IoT Cyber Security -- 6.3.1 Introduction to IoT Cyber Security -- 6.3.2 IoT Malware -- 6.4 Background of Machine Learning -- 6.4.1 Support Vector Machine (SVM) -- 6.4.2 Random Forest -- 6.4.3 Deep Neural Network (DNN) -- 6.5 Experiment -- 6.5.1 Training and Test Data -- 6.5.2 Baselines of the Machine Learning Models -- 6.6 Results and Discussion -- 6.6.1 Results -- 6.6.2 Discussion. , 6.7 Conclusion -- References -- Part IIIDeep Learning in the Medical Field -- 7 Survey on Deep Learning Techniques for Medical Imaging Application Area -- 7.1 Introduction -- 7.2 From Machine Learning to Deep Learning -- 7.3 Learning Algorithm -- 7.4 ANN -- 7.4.1 Activation Function in ANN -- 7.4.2 Training Process -- 7.5 DNN -- 7.5.1 Supervised Deep Learning -- 7.5.2 Unsupervised Learning -- 7.6 MRI Preprocessing -- 7.6.1 Inter-series Sorting -- 7.6.2 Registration -- 7.6.3 Normalization -- 7.6.4 Correction of the Bias Field -- 7.7 Deep Learning Applications in Medical Imagining -- 7.7.1 Classification -- 7.7.2 Detection -- 7.7.3 Segmentation -- 7.7.4 Registration -- 7.8 Conclusion -- References -- 8 Deep Learning Methods in Electroencephalography -- 8.1 Introduction -- 8.1.1 A Short Introduction to EEG -- 8.2 Literature Review -- 8.2.1 Public Datasets -- 8.2.2 Preprocessing Methods -- 8.2.3 Input Representation -- 8.2.4 Data Augmentation -- 8.2.5 Architectures -- 8.2.6 Features Visualization -- 8.2.7 Applications -- 8.3 Practical Example-Eriksen Flanker Task -- 8.3.1 Materials -- 8.4 Summary -- References -- Part IVDeep Learning in Systems Control -- 9 The Implementation and the Design of a Hybriddigital PI Control Strategy Based on MISO Adaptive Neural Network Fuzzy Inference System Models-A MIMO Centrifugal Chiller Case Study -- 9.1 Introduction -- 9.2 Centrifugal Chiller System Decomposition-Closed-Loop Simulations -- 9.3 MISO ARMAX and ANFIS Models of MIMO Centrifugal Chiller Plant -- 9.3.1 MISO ARMAX and ANFIS Evaporator Subsystem Models -- 9.3.2 MISO ARMAX and ANFIS Condenser Subsystem Models -- 9.4 Centrifugal Chiller PID Closed-Loop Control Strategies-Performance Analysis -- 9.5 Conclusions -- References -- 10 A Review of Deep Reinforcement Learning Algorithms and Comparative Results on Inverted Pendulum System -- 10.1 Introduction. , 10.2 Reinforcement Learning Background -- 10.2.1 Markov Decision Process -- 10.2.2 Deep-Q Learning -- 10.2.3 Double Deep-Q Learning -- 10.2.4 Double Dueling Deep-Q Learning -- 10.2.5 Reinforce -- 10.2.6 Asynchronous Deep Reinforcement Learning Methods -- 10.3 Inverted Pendulum Problem -- 10.4 Experimental Results -- 10.5 Conclusions -- References -- Part VDeep Learning in Feature Vector Processing -- 11 Stock Market Forecasting by Using Support Vector Machines -- 11.1 Introduction -- 11.2 Support Vector Machines -- 11.3 Determinants of Risk and Volatility in Stock Prices -- 11.4 Predictions of Stock Market Movements by Using SVM -- 11.4.1 Data Processing -- 11.4.2 The Proposed SVM Model -- 11.4.3 Feature Selection -- 11.5 Results and Conclusions -- References -- 12 An Experimental Exploration of Machine Deep Learning for Drone Conflict Prediction -- 12.1 Introduction -- 12.1.1 Airspace and Traffic Assumptions -- 12.1.2 Methodological Assumptions -- 12.2 A Brief Introduction to Artificial Neural Networks (ANNs) -- 12.3 Drone Test Scenarios and Traffic Samples -- 12.3.1 Experimental Design -- 12.3.2 ANN Design -- 12.3.3 Procedures -- 12.4 Results -- 12.4.1 Binary Classification Accuracy -- 12.4.2 Classification Sensitivity and Specificity -- 12.4.3 The Extreme Scenario -- 12.4.4 ROC Analysis -- 12.4.5 Summary of Results -- 12.5 Conclusions -- References -- 13 Deep Dense Neural Network for Early Prediction of Failure-Prone Students -- 13.1 Introduction -- 13.2 Literature Review -- 13.3 The Deep Dense Neural Network -- 13.4 Experimental Process and Results -- 13.5 Conclusions -- References -- Part VIEvaluation of Algorithm Performance -- 14 Non-parametric Performance Measurement with Artificial Neural Networks -- 14.1 Introduction -- 14.2 Data Envelopment Analysis -- 14.3 Artificial Neural Networks -- 14.4 Proposed Approach. , 14.4.1 Data Generation-Training and Testing Samples -- 14.4.2 ANN Architecture and Training Algorithm -- 14.5 Results -- 14.6 Conclusion -- References -- 15 A Comprehensive Survey on the Applications of Swarm Intelligence and Bio-Inspired Evolutionary Strategies -- 15.1 Introduction -- 15.2 Nature Inspired Intelligence -- 15.2.1 Swarm Intelligence -- 15.2.2 Algorithms Inspired by Organisms -- 15.3 Application Areas and Open Problems for NII -- 15.3.1 Applications of Swarm Intelligent Methods -- 15.3.2 Applications of Organisms-Inspired Algorithms -- 15.3.3 Comparison and Discussion -- 15.3.4 Are All These Algorithms Actually Needed? -- 15.4 Suggestions and Future Work -- References -- 16 Detecting Magnetic Field Levels Emitted by Tablet Computers via Clustering Algorithms -- 16.1 Introduction -- 16.2 Measurement of the Tablet Magnetic Field -- 16.2.1 Magnetic Field -- 16.2.2 Measuring Devices -- 16.2.3 TCO Standard -- 16.2.4 The Realized Experiment -- 16.2.5 A Typical Way of Working with the Tablet -- 16.3 Magnetic Field Clustering -- 16.3.1 K-Means Clustering -- 16.3.2 K-Medians Clustering -- 16.3.3 Self-Organizing Map Clustering -- 16.3.4 DBSCAN Clustering -- 16.3.5 Expectation-Maximization with Gaussian Mixture Models -- 16.3.6 Hierarchical Clustering -- 16.3.7 Fuzzy-C-Means Clustering -- 16.4 Evaluation of the Tablet User Exposure to ELF Magnetic Field -- 16.5 Results and Discussion -- 16.5.1 Measurement Results -- 16.5.2 Clustering Results -- 16.5.3 The foe Results Measurement -- 16.6 Conclusions -- References.