Logistic Regression, Random Forest, and Support Vector Machine.
PROJECT 03 / Classification · Healthcare
Healthcare Resource Modeling
A comparison of machine learning methods designed to help planners understand which cases may require resources and how confidently a model can support that decision.
DATASETHealthcare resource classification dataset
SIZESize not documented
SOURCECoursework dataset, public attribution pending
MEASURED EVIDENCE
What can be verified.
Metrics and outputs drawn from the project artifacts—not estimates added for presentation.
Principal component analysis was used to test whether correlated variables could be reduced.
Designed for resource planning exploration, not clinical decisions or patient diagnosis.
VALIDATION
Scope based evidence only: the case documents model comparison and responsible use boundaries. Sample size, class balance, and final benchmark scores were not preserved in the public project record.
Why this project matters
Healthcare teams often have limited staff, time, beds, or services and must decide where resources may be needed most. This project explored whether patterns in historical data could help organize that planning while keeping the model comparison transparent.
What was developed
The evaluation compared Logistic Regression, Random Forest, and Support Vector Machine models, then tested PCA to reduce overlapping information among correlated variables. Looking across several methods made it possible to examine tradeoffs instead of treating one algorithm as automatically best.
What it means
The result is a decision support prototype that demonstrates how classification can help planners identify patterns and prioritize review. It is not a clinical tool; its value is showing how model performance, explainability, and data quality shape responsible resource planning.
04 / WHERE THIS WORK APPLIES
From project to practical use.
Related concepts can support hospital capacity planning, patient outreach prioritization, public health programs, insurance operations, appointment scheduling, social service referrals, and nonprofit healthcare resource allocation.
05 / LIMITATION
What the project does not solve yet.
Performance may be affected by class imbalance, correlated variables, and limited representation across patient populations; the prototype is not a clinical decision system.
06 / NEXT ITERATION
Future advancement
Expand external validation, fairness and calibration testing, explainability, and prospective evaluation with healthcare stakeholders.
TECHNICAL TOOLKIT
Scikit-learnLogistic RegressionRandom ForestSVMPCA