Minimum Mismatch Modeling (3M) Hyperparameter Selection in Autoregressive Moving Average (ARMA) Modeling
Bibliographic record
Abstract
Hyperparameters of Autoregressive Moving Average (ARMA) modeling are the number of AR coefficients and the number of MA coefficients. The hyperparameter selection (HS) in ARMA modeling plays a critical role and can dominate the coefficient (parameter) estimation process. This work provides a novel method of HS estimation that works with the Conditional Least Square Estimator (CLSE), which is the most efficient ARMA parameter estimator. The proposed HS method focuses on a rational cost function in the form of mismatch modeling error. The error aims to capture the estimation difference between the true and unknown HS parameters and the competing hyperparameters. This error can be calculated using the available mean square error (MSE) in the parameter estimation step. The proposed method, denoted by the minimum mismatch modeling (3M) approach, has already shown superiority over other HS approaches in AR modeling. In AR modeling, the parameter estimator is based on the Yule-Walker method, which is a linear estimator, and the 3M calculation process using the available MSE has been provided for this modeling. However, in ARMA modeling the CLSE estimator is a nonlinear estimator, and one main challenge is to solve for calculation of the 3M using the MSE of CLSE. The method proposed here, denoted by 3M-CLSE, provides the steps to get to the desired 3M from the available CLSE MSE. It can be shown that the criteria of most used HS methods Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) are special cases of 3M-CLSE for particular choices of confidence and validation probabilities. The simulation results confirm the superiority of 3M-CLSE over the existing HS approaches in terms of HS accuracy, as well as in terms of modeling MSE error.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".