Swarm Intelligence and Its Applications in Biomedical Informatics

Cover
Half Title
Title Page
Copyright Page
Dedication
Table of Contents
Preface
Chapter 1 Introduction
	1.1 Background of Biomedical Informatics
		1.1.1 Methods in Biomedical Informatics
	1.2 Medical Aspects of Communicable and Noncommunicable Diseases
		1.2.1 Statistical Report for Risk Factors of NCD
		1.2.2 Epidemiology of Non-Communicable Disease
		1.2.3 Purpose of Epidemiological Study and Assumptions
	1.3 Biomedical Informatics
		1.3.1 IT-Oriented Definitions
		1.3.2 Computation – A Tool for Informatics
		1.3.3 Benefits and Applications of Biomedical Informatics
		1.3.4 Structural Bioinformatics and Proteome-Based Technology
		1.3.5 Bio-programming and Bioinformatics Software
		1.3.6 Disease Patterns and Treatment Analysis
		1.3.7 Scientific Demand Analysis and Outcomes
		1.3.8 Applications
	1.4 Application of Big Data in Biomedical Informatics
		1.4.1 Big Data – An Introduction
		1.4.2 The Four Vs of Big Data
		1.4.3 Importance of Big Data in Clinical Informatics
		1.4.4 Visual Perception with Medical Data
	1.5 Analytical Techniques on Medical Data
		1.5.1 Clinical Prediction Models
	References
Chapter 2 Statistical Methods and Swarm Intelligence Techniques in Clinical Data Prediction
	2.1 Swarm Intelligence an Overview
	2.2 Swarm Intelligence in Data Prediction and Decision-Making
		2.2.1 Characteristics of Swarm Intelligence
		2.2.2 Swarm Intelligence Techniques
	2.3 Statistical Approaches in Medical Data Analytics
		2.3.1 Regression Analysis
		2.3.2 Linear Discriminant Analysis
		2.3.3 Correlation Analysis
	2.4 Diagnostic Evaluation by Receiver Operating Characteristic Curve
	2.5 Summary of Swarm Intelligence in Biomedical Informatics
	References
Chapter 3 Data Classification by Decision Trees – An Illustration
	3.1 Introduction
	3.2 Variants of Decision Trees
	3.3 Decision Trees Splitting Measures
		3.3.1 Information Gain
		3.3.2 Gain Ratio
		3.3.3 Gini Index
	3.4 Example Illustration
		3.4.1 Gain Ratio
		3.4.2 Gini Index
	3.5 Applications
	References
Chapter 4 Predictive Analytics Using Ant Colony Optimization with Decision Trees for Medical Data
	4.1 Data Pre-Processing
	4.2 ACO – Decision Trees
		4.2.1 ACO Pheromone Initialization
		4.2.2 Generate Population
		4.2.3 Select Features According to State Transition Rule
		4.2.4 Pheromone Updating Using Update Rule
		4.2.5 Sort Features According to Pheromone Values
		4.2.6 Build Feature Set with Top Pheromone Values
		4.2.7 ACO – Decision Tree Algorithm Working Procedure
		4.2.8 Cross-Validation
		4.2.9 Evaluate Fitness for Selected Features Using Decision Trees
	4.3 Experimentation of the Developed Model Over Various Medical Datasets
		4.3.1 Real-World Datasets
		4.3.2 Benchmark Datasets
	4.4 Analysis of Time Complexity
	4.5 Determination of Risk Correlation with Selected Features
	4.6 Summary
	References
Chapter 5 Predictive Analytics Using Bee-Based Harmony Search with Decision Trees for Medical Data
	5.1 A Review on Clinical Data Analysis
	5.2 Data Collection and Pre-Processing
	5.3 Pathogenesis of Non-Communicable Diseases
	5.4 Proposed Model Using BHS – Decision Trees
		5.4.1 Harmony Memory Initialization
		5.4.2 Memory Concern in Harmony Search
		5.4.3 Classification Over Selected Features Using Decision Trees
		5.4.4 Selection of Splitting Criterion
	5.5 Experimentation of the Observed Model Over Various Medical Datasets
		5.5.1 Performance Metrics
		5.5.2 Confusion Matrix
		5.5.3 Experimental Analysis with Linear Model Development for Type II Diabetic Dataset
	5.6 Summary
	References
Chapter 6 Predictive Analytics Using Particle Swarm Optimization with Decision Trees for Type II Diabetes
	6.1 A Review on Type II Diabetes and its Implications
	6.2 Rationale Behind the Proposed Approach
	6.3 Proposed Method Using PSO-J48
		6.3.1 Algorithm Description
		6.3.2 Proposed Workflow
	6.4 Experimental Results and Discussion
		6.4.1 Learning Population
		6.4.2 Attribute Description and Measurement
		6.4.3 PSO Operating Parameters and Learning Factors
		6.4.4 Modified Self-adaptive Inertial Weight with Convergence Logic
		6.4.5 Metrics for Evaluation
		6.4.6 Cross-Validation and Results Obtained
	6.5 Formulating the Association Among the Risk Factors by Mathematical Model Using Fisher's Linear Discriminant Analysis for Type II Diabetic Prediction
		6.5.1 Association of Type II Diabetic Risk with Attribute Measurements
		6.5.2 Comparison of the Proposed Model with the Existing Approaches
		6.5.3 Test Interpretation
		6.5.4 Test Evaluation
		6.5.5 ROC Analysis
	6.6 Performance of the Proposed Approaches for Real-World Datasets
		6.6.1 Summary
	References
Chapter 7 Case-Based Analysis in Medical Informatics
	7.1 Diabetic Primary
	Case Study Report
		7.1.1 Challenges Faced in Type II Diabetes
		7.1.2 Epidemiology
		7.1.3 Aetiology
		7.1.4 Making the Diagnosis
		7.1.5 Prognosis
		7.1.6 Factors to be Considered for Blood Test
		7.1.7 Medications Available for T2DM
		7.1.8 Algorithms Used to Analyze Diabetics
		7.1.9 Telemonitoring Evidence-based Analysis for Type 2 Diabetics
	References
Chapter 8 Intelligent Optimization Unit
	8.1 Proposed Framework
		8.1.1 Min-Max Normalization
		8.2.2 Z-Score Normalization
		8.3.3 Normalization by Decimal Scaling
	8.2 Dataset Description
		8.2.1 Benchmark Data and its Description
		8.2.2 Real-World Data Collected From Hospitals
	References
Chapter 9 Conclusion
	9.1 Summary of the Research Work and its Significant Outcomes
		9.1.1 Research Findings
		9.1.2 Limitations
	9.2 Future Work
Index