Machine Learning and Predicting Clinical Outcomes [HBNI Th259]

Chandrani Kumari

dc.contributor.advisor
dc.contributor.author	Chandrani Kumari
dc.date.accessioned	2025-04-11T08:32:04Z
dc.date.available	2025-04-11T08:32:04Z
dc.date.issued	2025
dc.date.submitted	2025-02-28
dc.identifier.uri	https://dspace.imsc.res.in/xmlui/handle/123456789/895
dc.description.tableofcontents	1. Introduction – 15 1.1 AI in the Healthcare Sector – 16 1.2 Designing Machine Learning Models – 19 • 1.2.1 Data Preprocessing – 19 • 1.2.2 Model Development – 23 • 1.2.3 Evaluation Metrics – 23 1.3 Progress, Challenges, and Opportunities for AI in Healthcare – 28 • 1.3.1 Challenges for AI in Healthcare – 28 • 1.3.2 Opportunities for AI in Healthcare – 31 2. Machine Learning Algorithms – 33 2.1 Supervised Learning – 34 • 2.1.1 k-Nearest Neighbors (k-NN) – 34 • 2.1.2 Linear Regression – 35 • 2.1.3 Logistic Regression – 35 • 2.1.4 Naive Bayes – 36 • 2.1.5 Decision Tree – 38 • 2.1.6 Support Vector Machine – 39 2.2 Unsupervised Learning – 39 • 2.2.1 K-Means – 40 • 2.2.2 Hierarchical Clustering – 41 • 2.2.3 DBSCAN – 42 • 2.2.4 Gaussian Mixture Model (GMM) – 43 • 2.2.5 Principal Component Analysis (PCA) – 44 • 2.2.6 Other Dimensionality Reduction Techniques – 45 2.3 Ensemble Models – 45 • 2.3.1 Boosting – 46 • 2.3.2 Bagging – 47 2.4 Neural Networks – 48 2.5 Reinforcement Learning – 49 2.6 Clinical Study Design and Evaluation Metrics – 50 3. A Predictive Model of a Growing Fetus: The Gompertz Model – 52 3.1 Introduction – 52 3.2 Materials and Methods – 56 • 3.2.1 Study Design and Participants – 56 • 3.2.2 Published Growth Standards for Fetal Biometry – 58 • 3.2.3 Gompertz Growth Model – 59 • 3.2.4 Quality of Fit – 61 • 3.2.5 Predicting Birth Weight of Fetus – 61 • 3.2.6 Validation Cohort – 63 3.3 Results – 64 • 3.3.1 Gompertz Formula Fits Biometry Measurements (Seethapathy Cohort) – 64 • 3.3.2 Neonatal Complications and Deviations from Gompertz Growth – 65 • 3.3.3 Predicting Final Biometry and Individual Birth Weights – 66 • 3.3.4 Model Performance on Validation Cohort – 70 • 3.3.5 Gompertz Equation and Fetal Growth Standards – 71 3.4 Conclusion – 76 4. MMM and MMMSynth: Clustering and Synthetic Data Generation – 79 4.1 Introduction – 79 • 4.1.1 Existing Clustering Methods – 80 • 4.1.2 Overview of the MMM Algorithm – 81 • 4.1.3 MMMSynth: Generating Synthetic Tabular Data – 82 4.2 Methods – 83 • 4.2.1 MMM: Clustering of Heterogeneous Data – 83 • 4.2.2 Discrete Data, Dirichlet Prior – 83 • 4.2.3 Continuous Data, Normal-Gamma Prior – 85 • 4.2.4 Optimizing Clustering Likelihood (Expectation Maximization) – 87 • 4.2.5 Identifying the Correct K: Marginal Likelihood – 89 • 4.2.6 Assessment of Benchmarking – 95 • 4.2.7 MMMSynth: Generating Synthetic Data with MMM – 96 • 4.2.8 Benchmarks: Real Datasets Used – 97 4.3 Results – 99 • 4.3.1 Clustering Algorithm Performance – 99 4.4 Conclusion – 106 5. Other Machine Learning Projects – 109 5.1 Assessing Vasoplegia Grouping using ML – 109 5.2 DREAM Preterm Birth Prediction Challenge (Transcriptomics) – 116 5.3 RSNA-MICCAI Brain Tumor Radiogenomic Classification – 117 6. Conclusion – 119	en_US
dc.publisher.publisher	IMSc
dc.publisher.publisher	The Institute of Mathematical Sciences
dc.relation.isbasedon		en_US
dc.subject	Predictive Analytics	en_US
dc.subject	Artificial intelligence
dc.subject	Decision support systems
dc.subject	Machine learning
dc.title	Machine Learning and Predicting Clinical Outcomes [HBNI Th259]	en_US
dc.type.degree	Ph.D	en_US
dc.date.updated	2025
dc.date.updated	2025
dc.type.institution	HBNI	en_US
dc.description.advisor	Rahul Siddharthan
dc.description.pages	133p.	en_US
dc.type.mainsub	Computational Biology	en_US
dc.type.hbnibos	Life Sciences	en_US