Information Criterion (AIC and BIC) are used to compare competing r…

Question

Information Criterion (AIC and BIC) are used to compare competing regression models. Which statement correctly differentiates AIC from BIC for model selection?

Accepted Answer

BIC penalizes model complexity more severely than AIC for large sample sizes because its penalty term grows with log(n), tending toward more parsimonious models — Here's the deal — b is correct because BIC = -2log(L) + klog(n) while AIC = -2log(L) + 2k. For large n (log(n) > 2, i.e., n > 7), BIC penalizes parameters more heavily and tends to select simpler models.

Answer

AIC penalizes model complexity more severely than BIC for large sample sizes, favoring parsimonious models

Answer

AIC and BIC produce identical model rankings when the sample size exceeds 100 observations

Answer

BIC is preferred when prediction accuracy is the primary goal; AIC is preferred when identifying the true data generating model

Information Criterion (AIC and BIC) are used to compare competing regression models. Which statement correctly differentiates AIC from BIC for model selection?

Short Explanation and Infographic

Full Explanation