econometron.utils.estimation
Overview
The econometron.utils.estimation submodule provides essential utilities for parameter estimation in econometric, statistical, and time series models. It includes robust algorithms for maximum likelihood, Bayesian, and classical estimation, as well as functions for diagnostics, results formatting, and model comparison. The module is designed for flexibility, reproducibility, and integration with the broader Econometron library.
Theoretical Background
Parameter estimation is central to econometric modeling, enabling inference, prediction, and hypothesis testing. This module supports:
- Maximum Likelihood Estimation (MLE) via metaheuristic and classical algorithms
- Bayesian estimation techniques
- Ordinary Least Squares (OLS) for linear models
- Diagnostics
Module Contents
| Function/Class | Description |
|---|---|
genetic_algorithm_kalman | Genetic Algorithm for MLE with Kalman filter |
simulated_annealing_kalman | Simulated Annealing for MLE with Kalman filter |
ols_estimator | Multivariate Ordinary Least Squares regression |
compute_stats | Compute model diagnostics and statistics |
make_prior_func | Create prior functions for Bayesian estimation |
rwm_kalman | Random Walk Metropolis for Bayesian Kalman filtering |
compute_prior | Compute prior probabilities for Bayesian models |
Notes
- All estimation functions are designed for integration with Econometron models and utilities.
- Results formatting and diagnostics are standardized for reproducibility and transparency.
- Bayesian and MLE routines support flexible priors and optimization settings.
References
- Greene, W. H. (2018). Econometric Analysis. Pearson.
- Hamilton, J. D. (1994). Time Series Analysis. Princeton University Press.
- Gelman, A., et al. (2013). Bayesian Data Analysis. CRC Press.
- Wikipedia: Maximum Likelihood Estimation
- Wikipedia: Bayesian Inference