Support Vector Machine Modeling & Graph Theory Metrics for Disease Classification

on YouTube ↗15m

" summary

Combines BRFSS survey data with a simulated small‑world social network (rewired lattice created in R) and runs SVMs in SAS alongside logistic regression; key predictors are age, education, BMI, hypertension, cholesterol and a centrality measure that reflects household size. Logistic regression yields ROC 81.8 vs 81.7 for models with and without graph metrics, while SVM with a polynomial kernel gives ROC 72.9 and 75.6, so SVM underperforms here though graph metrics slightly help; code and a customized kernel path are planned. Behavioral Risk Factor Surveillance System

AI-written summary · about

" chapters (12)

1. 0:00Intro
2. 0:51OVERVIEW
3. 2:04Support Vector Machine
4. 4:02SVM application for disease classification
5. 4:51Graph Theory application for disease classification
6. 7:43Social Network Simulation
7. 9:07PROC OPTGRAPH Code
8. 10:20Logistic Regression Results
9. 11:22PROC HPSVM Code
10. 11:58SVM Penalty, C, Selection
11. 12:54Model Comparison (with graph metrics)
12. 13:14Future Research

" related talks

• 20mML applications in healthcare - Saudi Arabian hospital case studyAmibath Saxena · DSC Europe 2022
• 38mMixed and Covariate-Dependent Graphical ModelsLiza Levina · Workshop on Succinct Data Representations and Applications (Simons Institute Big Data 2013)
• 17mLeveraging Knowledge Graph and Imaging to Predict Gene–Disease Links in Cardiovascular DiseasesKhaled Rjoob · NODES 2025
• 50mThe Shape of Data: Machine Learning and TopologyKaisa Taipale · NDC Minnesota 2019
• 35mOn Machine Learning for Prediction and Prioritization in the Allocation of Scarce Societal ResourcesSanmay Das · Bridging Prediction and Intervention Problems in Social Systems (Simons Institute, 2026)
• 58mIntroduction to Biological Network Analysis I: Network Basics and PropertiesDonna Slonim · Algorithmic Challenges in Genomics Boot Camp
• 45mUsing Support Vector Machines in Scikit Learn to discover genetic aetiologiesTim Vivian Griffiths · PyData London 2016
• 26mMachine Learning for Personalised HealthDanielle Belgrave · Strange Loop 2018