Akaike information criterion
Criterio de información de Akaike
- Date
- 1973 · confidence high
- Place
- Tokyo, Japan
- Credited to
- Hirotugu Akaike
- Field
- Knowledge & symbols
What was discovered
Statistical criterion estimating the relative quality of models for a given data set, balancing model fit with complexity; formulated by Hirotugu Akaike.
What it led to
- Statistical model selection and prevention of overfitting
- Basis for other information criteria (BIC, corrected AIC)
- Applications in econometrics, ecology and machine learning