Stepwise regression

Last updated

In statistics, stepwise regression is a method of fitting regression models in which the choice of predictive variables is carried out by an automatic procedure. [1] [2] [3] [4] In each step, a variable is considered for addition to or subtraction from the set of explanatory variables based on some prespecified criterion. Usually, this takes the form of a forward, backward, or combined sequence of F-tests or t-tests.

Contents

The frequent practice of fitting the final selected model followed by reporting estimates and confidence intervals without adjusting them to take the model building process into account has led to calls to stop using stepwise model building altogether [5] [6] or to at least make sure model uncertainty is correctly reflected by using prespecified, automatic criteria together with more complex standard error estimates that remain unbiased. [7] [8]

In this example from engineering, necessity and sufficiency are usually determined by F-tests. For additional consideration, when planning an experiment, computer simulation, or scientific survey to collect data for this model, one must keep in mind the number of parameters, P, to estimate and adjust the sample size accordingly. For K variables, P = 1(Start) + K(Stage I) + (K - K)/2(Stage II) + 3K(Stage III) = 0.5K + 3.5K + 1. For K < 17, an efficient design of experiments exists for this type of model, a Box-Behnken design, augmented with positive and negative axial points of length min(2, (int(1.5 + K/4)) ), plus point(s) at the origin. There are more efficient designs, requiring fewer runs, even for K > 16. Stepwise.jpg
In this example from engineering, necessity and sufficiency are usually determined by F-tests. For additional consideration, when planning an experiment, computer simulation, or scientific survey to collect data for this model, one must keep in mind the number of parameters, P, to estimate and adjust the sample size accordingly. For K variables, P = 1(Start) + K(Stage I) + (K − K)/2(Stage II) + 3K(Stage III) = 0.5K + 3.5K + 1. For K < 17, an efficient design of experiments exists for this type of model, a Box–Behnken design, augmented with positive and negative axial points of length min(2, (int(1.5 + K/4)) ), plus point(s) at the origin. There are more efficient designs, requiring fewer runs, even for K > 16.

Main approaches

The main approaches for stepwise regression are:

Alternatives

A widely used algorithm was first proposed by Efroymson (1960). [10] This is an automatic procedure for statistical model selection in cases where there is a large number of potential explanatory variables, and no underlying theory on which to base the model selection. The procedure is used primarily in regression analysis, though the basic approach is applicable in many forms of model selection. This is a variation on forward selection. At each stage in the process, after a new variable is added, a test is made to check if some variables can be deleted without appreciably increasing the residual sum of squares (RSS). The procedure terminates when the measure is (locally) maximized, or when the available improvement falls below some critical value.

One of the main issues with stepwise regression is that it searches a large space of possible models. Hence it is prone to overfitting the data. In other words, stepwise regression will often fit much better in sample than it does on new out-of-sample data. Extreme cases have been noted where models have achieved statistical significance working on random numbers. [11] This problem can be mitigated if the criterion for adding (or deleting) a variable is stiff enough. The key line in the sand is at what can be thought of as the Bonferroni point: namely how significant the best spurious variable should be based on chance alone. On a t-statistic scale, this occurs at about , where p is the number of predictors. Unfortunately, this means that many variables which actually carry signal will not be included. This fence turns out to be the right trade-off between over-fitting and missing signal. If we look at the risk of different cutoffs, then using this bound will be within a factor of the best possible risk. Any other cutoff will end up having a larger such risk inflation. [12] [13]

Model accuracy

A way to test for errors in models created by step-wise regression, is to not rely on the model's F-statistic, significance, or multiple R, but instead assess the model against a set of data that was not used to create the model. [14] This is often done by building a model based on a sample of the dataset available (e.g., 70%) – the “training set” – and use the remainder of the dataset (e.g., 30%) as a validation set to assess the accuracy of the model. Accuracy is then often measured as the actual standard error (SE), MAPE (Mean absolute percentage error), or mean error between the predicted value and the actual value in the hold-out sample. [15] This method is particularly valuable when data are collected in different settings (e.g., different times, social vs. solitary situations) or when models are assumed to be generalizable.

Criticism

Stepwise regression procedures are used in data mining, but are controversial. Several points of criticism have been made.

Such criticisms, based upon limitations of the relationship between a model and procedure and data set used to fit it, are usually addressed by verifying the model on an independent data set, as in the PRESS procedure.

Critics regard the procedure as a paradigmatic example of data dredging, intense computation often being an inadequate substitute for subject area expertise. Additionally, the results of stepwise regression are often used incorrectly without adjusting them for the occurrence of model selection. Especially the practice of fitting the final selected model as if no model selection had taken place and reporting of estimates and confidence intervals as if least-squares theory were valid for them, has been described as a scandal. [7] Widespread incorrect usage and the availability of alternatives such as ensemble learning, leaving all variables in the model, or using expert judgement to identify relevant variables have led to calls to totally avoid stepwise model selection. [5]

See also

Related Research Articles

<span class="mw-page-title-main">Overfitting</span> Flaw in mathematical modelling

In mathematical modeling, overfitting is "the production of an analysis that corresponds too closely or exactly to a particular set of data, and may therefore fail to fit to additional data or predict future observations reliably". An overfitted model is a mathematical model that contains more parameters than can be justified by the data. In a mathematical sense, these parameters represent the degree of a polynomial. The essence of overfitting is to have unknowingly extracted some of the residual variation as if that variation represented underlying model structure.

<span class="mw-page-title-main">Cross-validation (statistics)</span> Statistical model validation technique

Cross-validation, sometimes called rotation estimation or out-of-sample testing, is any of various similar model validation techniques for assessing how the results of a statistical analysis will generalize to an independent data set. Cross-validation includes resampling and sample splitting methods that use different portions of the data to test and train a model on different iterations. It is often used in settings where the goal is prediction, and one wants to estimate how accurately a predictive model will perform in practice. It can also be used to assess the quality of a fitted model and the stability of its parameters.

<span class="mw-page-title-main">Data dredging</span> Misuse of data analysis

Data dredging is the misuse of data analysis to find patterns in data that can be presented as statistically significant, thus dramatically increasing and understating the risk of false positives. This is done by performing many statistical tests on the data and only reporting those that come back with significant results.

<span class="mw-page-title-main">Coefficient of determination</span> Indicator for how well data points fit a line or curve

In statistics, the coefficient of determination, denoted R2 or r2 and pronounced "R squared", is the proportion of the variation in the dependent variable that is predictable from the independent variable(s).

<span class="mw-page-title-main">Structural equation modeling</span> Form of causal modeling that fit networks of constructs to data

Structural equation modeling (SEM) is a diverse set of methods used by scientists doing both observational and experimental research. SEM is used mostly in the social and behavioral sciences but it is also used in epidemiology, business, and other fields. A definition of SEM is difficult without reference to technical language, but a good starting place is the name itself.

In statistics, a generalized additive model (GAM) is a generalized linear model in which the linear response variable depends linearly on unknown smooth functions of some predictor variables, and interest focuses on inference about these smooth functions.

Omnibus tests are a kind of statistical test. They test whether the explained variance in a set of data is significantly greater than the unexplained variance, overall. One example is the F-test in the analysis of variance. There can be legitimate significant effects within a model even if the omnibus test is not significant. For instance, in a model with two independent variables, if only one variable exerts a significant effect on the dependent variable and the other does not, then the omnibus test may be non-significant. This fact does not affect the conclusions that may be drawn from the one significant variable. In order to test effects within an omnibus test, researchers often use contrasts.

In statistics, Mallows's, named for Colin Lingwood Mallows, is used to assess the fit of a regression model that has been estimated using ordinary least squares. It is applied in the context of model selection, where a number of predictor variables are available for predicting some outcome, and the goal is to find the best model involving a subset of these predictors. A small value of means that the model is relatively precise.

In statistics, model specification is part of the process of building a statistical model: specification consists of selecting an appropriate functional form for the model and choosing which variables to include. For example, given personal income together with years of schooling and on-the-job experience , we might specify a functional relationship as follows:

In statistics, regression validation is the process of deciding whether the numerical results quantifying hypothesized relationships between variables, obtained from regression analysis, are acceptable as descriptions of the data. The validation process can involve analyzing the goodness of fit of the regression, analyzing whether the regression residuals are random, and checking whether the model's predictive performance deteriorates substantially when applied to data that were not used in model estimation.

<span class="mw-page-title-main">Least-angle regression</span>

In statistics, least-angle regression (LARS) is an algorithm for fitting linear regression models to high-dimensional data, developed by Bradley Efron, Trevor Hastie, Iain Johnstone and Robert Tibshirani.

In statistics, multivariate adaptive regression splines (MARS) is a form of regression analysis introduced by Jerome H. Friedman in 1991. It is a non-parametric regression technique and can be seen as an extension of linear models that automatically models nonlinearities and interactions between variables.

Psychometric software refers to specialized programs used for the psychometric analysis of data that was obtained from tests, questionnaires, polls or inventories that measure latent psychoeducational variables. Although some psychometric analysis can be conducted using general statistical software like SPSS, most require dedicated tools designed specifically for psychometric purposes.

The following outline is provided as an overview of and topical guide to regression analysis:

<span class="mw-page-title-main">Exploratory factor analysis</span> Statistical method in psychology

In multivariate statistics, exploratory factor analysis (EFA) is a statistical method used to uncover the underlying structure of a relatively large set of variables. EFA is a technique within factor analysis whose overarching goal is to identify the underlying relationships between measured variables. It is commonly used by researchers when developing a scale and serves to identify a set of latent constructs underlying a battery of measured variables. It should be used when the researcher has no a priori hypothesis about factors or patterns of measured variables. Measured variables are any one of several attributes of people that may be observed and measured. Examples of measured variables could be the physical height, weight, and pulse rate of a human being. Usually, researchers would have a large number of measured variables, which are assumed to be related to a smaller number of "unobserved" factors. Researchers must carefully consider the number of measured variables to include in the analysis. EFA procedures are more accurate when each factor is represented by multiple measured variables in the analysis.

In statistics, model validation is the task of evaluating whether a chosen statistical model is appropriate or not. Oftentimes in statistical inference, inferences from models that appear to fit their data may be flukes, resulting in a misunderstanding by researchers of the actual relevance of their model. To combat this, model validation is used to test whether a statistical model can hold up to permutations in the data. This topic is not to be confused with the closely related task of model selection, the process of discriminating between multiple candidate models: model validation does not concern so much the conceptual design of models as it tests only the consistency between a chosen model and its stated outputs.

In statistics, the predicted residual error sum of squares (PRESS) is a form of cross-validation used in regression analysis to provide a summary measure of the fit of a model to a sample of observations that were not themselves used to estimate the model. It is calculated as the sum of squares of the prediction residuals for those observations. Specifically, the PRESS statistic is an exhaustive form of cross-validation, as it tests all the possible ways that the original data can be divided into a training and a validation set.

In statistics, the one in ten rule is a rule of thumb for how many predictor parameters can be estimated from data when doing regression analysis while keeping the risk of overfitting and finding spurious correlations low. The rule states that one predictive variable can be studied for every ten events. For logistic regression the number of events is given by the size of the smallest of the outcome categories, and for survival analysis it is given by the number of uncensored events.

In statistics, linear regression is a statistical model which estimates the linear relationship between a scalar response and one or more explanatory variables. The case of one explanatory variable is called simple linear regression; for more than one, the process is called multiple linear regression. This term is distinct from multivariate linear regression, where multiple correlated dependent variables are predicted, rather than a single scalar variable. If the explanatory variables are measured with error then errors-in-variables models are required, also known as measurement error models.

References

  1. ↑ Efroymson, M. A. (1960) "Multiple regression analysis," Mathematical Methods for Digital Computers, Ralston A. and Wilf, H. S., (eds.), Wiley, New York.
  2. ↑ Hocking, R. R. (1976) "The Analysis and Selection of Variables in Linear Regression," Biometrics, 32.
  3. ↑ Draper, N. and Smith, H. (1981) Applied Regression Analysis, 2d Edition, New York: John Wiley & Sons, Inc.
  4. ↑ SAS Institute Inc. (1989) SAS/STAT User's Guide, Version 6, Fourth Edition, Volume 2, Cary, NC: SAS Institute Inc.
  5. 1 2 Flom, P. L. and Cassell, D. L. (2007) "Stopping stepwise: Why stepwise and similar selection methods are bad, and what you should use," NESUG 2007.
  6. ↑ Harrell, F. E. (2001) "Regression modeling strategies: With applications to linear models, logistic regression, and survival analysis," Springer-Verlag, New York.
  7. 1 2 Chatfield, C. (1995) "Model uncertainty, data mining and statistical inference," J. R. Statist. Soc. A 158, Part 3, pp. 419–466.
  8. ↑ Efron, B. and Tibshirani, R. J. (1998) "An introduction to the bootstrap," Chapman & Hall/CRC
  9. ↑ Box–Behnken designs from a handbook on engineering statistics at NIST
  10. ↑ Efroymson, MA (1960) "Multiple regression analysis." In Ralston, A. and Wilf, HS, editors, Mathematical Methods for Digital Computers. Wiley.
  11. ↑ Knecht, WR. (2005). Pilot willingness to take off into marginal weather, Part II: Antecedent overfitting with forward stepwise logistic regression. (Technical Report DOT/FAA/AM-O5/15). Federal Aviation Administration
  12. ↑ Foster, Dean P., & George, Edward I. (1994). The Risk Inflation Criterion for Multiple Regression. Annals of Statistics, 22(4). 1947–1975. doi : 10.1214/aos/1176325766
  13. ↑ Donoho, David L., & Johnstone, Jain M. (1994). Ideal spatial adaptation by wavelet shrinkage. Biometrika, 81(3):425–455. doi : 10.1093/biomet/81.3.425
  14. ↑ Mark, Jonathan, & Goldberg, Michael A. (2001). Multiple regression analysis and mass assessment: A review of the issues. The Appraisal Journal, Jan., 89–109.
  15. ↑ Mayers, J.H., & Forgy, E.W. (1963). The Development of numerical credit evaluation systems. Journal of the American Statistical Association, 58(303; Sept), 799–806.
  16. ↑ Rencher, A. C., & Pun, F. C. (1980). Inflation of R² in Best Subset Regression. Technometrics, 22, 49–54.
  17. ↑ Copas, J.B. (1983). Regression, prediction and shrinkage. J. Roy. Statist. Soc. Series B, 45, 311–354.
  18. ↑ Wilkinson, L., & Dallal, G.E. (1981). Tests of significance in forward selection regression with an F-to enter stopping rule. Technometrics, 23, 377–380.
  19. ↑ Hurvich, C. M. and C. L. Tsai. 1990. The impact of model selection on inference in linear regression. American Statistician 44: 214–217.
  20. ↑ Roecker, Ellen B. (1991). Prediction error and its estimation for subset—selected models. Technometrics, 33, 459–468.
This page is based on this Wikipedia article
Text is available under the CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.