Logistic Regression

Linear Regression - Code

Parameter Tuning - Linear Models

• Linear Regression: predict sth. that is continuous

Lineare Regression vs Logistic Regression

Die lineare Regression wird für kontinuierliche Zielwerte verwendet, während die logistische Regression für Klassifikationsprobleme (meist binär) eingesetzt wird.

→ gute Erklärung hier

Checklist & Evaluation

Linear regression trust checklist:

Make sure the model is statistically significant before extracting any meaningful explanation!

• Residuals? ✅
• R-squared? ✅
• P-values? ✅
• Confidence intervals? ✅

Evaluation metrics for regression problems → Link

<aside> 💡 In general its better to work with Ridge models since they are much more robust against outlier and typically beats linear regression. See Regularization - Ridge (L2) Regression. Details in this calmcode.

</aside>