Medical Machine Learning & Prescriptive Analytics Quiz

Time Remaining:
120:00

šŸ„ Business Problem & Medical Context

Hospital Scenario: You are a data scientist at Metro General Hospital, a 500-bed facility that recently implemented three different treatment protocols (A, B, and C) for a common medical condition. The hospital leadership wants to optimize patient outcomes, reduce readmission rates, and improve operational efficiency.

Objective: Use machine learning and prescriptive analytics to:

Clinical Importance: With healthcare costs rising and patient outcomes under scrutiny, data-driven approaches to treatment selection and risk stratification are crucial for modern healthcare delivery.

šŸ“Š Dataset Description

Source: 500 anonymized patient records from Metro General Hospital

Variables:

Note: This synthetic dataset includes realistic medical relationships, such as Treatment B being more effective for older patients and higher readmission risk correlating with poorer outcomes.

Student Information

Part 1: Data Insights & Exploratory Analysis (5 Questions)

Analyze the dataset using descriptive statistics, correlations, and visualizations

Question 1: Overall Recovery Rate

Based on the medical dataset provided, what is the overall recovery rate across all 500 patients?

Question 2: Treatment Effectiveness

Which treatment type has the lowest average readmission risk score, indicating the safest treatment option?

Question 3: Age-Based Treatment Analysis

For patients over 70 years old, which treatment shows the highest recovery rate?

Question 4: Correlation Analysis

What is the correlation between patient age and hospital stay duration?

Question 5: High-Risk Patient Analysis

Among patients with blood pressure >140 AND cholesterol >200, what percentage received Treatment B?

Part 2: Machine Learning Model Application (5 Questions)

Train and evaluate classification models to predict patient recovery outcomes

Question 6: Model Performance

After training a logistic regression model to predict patient recovery outcomes, what is the expected model accuracy?

Question 7: Feature Importance

Which feature has the highest absolute coefficient (strongest predictive power) in the logistic regression model?

Question 8: Model Precision

What is the model's precision for predicting positive outcomes (recovery = 1)?

Question 9: High-Risk Predictions

When applying the trained model to patients with Readmission_Risk_Score > 80, what percentage are predicted to recover?

Question 10: Treatment Effect in Model

Based on the logistic regression coefficients, which statement about Treatment B is most accurate?

Part 3: Prescriptive Methodologies & Actionable Insights (10 Questions)

Apply SHAP values, ATE analysis, and clustering for clinical decision-making

Question 11: SHAP Analysis

Using SHAP values, which feature contributes most to positive predictions (recovery) for the highest-risk patients (top 20% by readmission risk)?

Question 12: Average Treatment Effect

What is the Average Treatment Effect (ATE) of Treatment B compared to Treatment A for patients over 50?

Question 13: K-means Clustering

When performing K-means clustering (k=3) on patient characteristics, which cluster has the highest average readmission risk score?

Question 14: Personalized Treatment

Based on SHAP analysis, for a 45-year-old patient with high cholesterol, which intervention would most improve their predicted outcome?

Question 15: Resource Allocation

According to clustering analysis, which patient cluster would benefit most from targeted intervention programs?

Question 16: Population-Level Impact

Based on your ATE analysis, what treatment protocol change would have the greatest population-level impact on hospital recovery rates?

Question 17: Clinical Decision Threshold

Using SHAP values to inform clinical decision-making, at what readmission risk score threshold should the hospital implement immediate intensive intervention protocols?

Question 18: Treatment Selection Criteria

For personalized treatment selection, which patient characteristic combination most strongly indicates Treatment B would be optimal?

Question 19: Cost-Effectiveness

Based on your model's feature importance, which single operational change would most cost-effectively improve patient outcomes hospital-wide?

Question 20: Triage Strategy

To optimize resource allocation, which patient triage strategy should the hospital implement at admission?

Quiz Results

Your score: 0/20