Statistical Learning from a Regression Perspective

By H. Carmichael. 2008

313 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Fairness in Criminal Justice Risk Assessments: The State of the Art (2017, methodology): “:3) ð L ; S Þ when conditional use accuracy is the same for both protected group classes (Berk 2016b).”
Handbook of Big Data (2016, background): “This, along with lack with the capability to perform statistical inference and other concerns, led Berk to write in 2008 [1] that shrinkage estimators such as the LASSO seem to primarily be niche players.”
Forecasting murder within a population of probationers and parolees: a high stakes application of statistical learning (2009, background): “Among the many obstacles to arriving at some overall evaluation is a general failure to quantify forecasting accuracy using data that were not previously employed in the process by which the forecasts were constructed (Berk, 2008a).”
Forecasting Domestic Violence: A Machine Learning Approach to Help Inform Arraignment Decisions (2016, background): “As a mathematical matter, it is only the relative costs that count (Berk 2008:Sec.”
What You Can and Can’t Properly Do with Regression (2010, background): “Among the more popular approaches are multivariate statistics (Anderson 1958; Gifi 1990), exploratory data analysis (Tukey 1977; Diaconis 1985), and more recently, dynamic graphics (Cook and Swayne 2007) and Machine/Statistical Learning (Hastie et al. 2009; Berk 2008).”
The Role of Race in Forecasts of Violent Crime (2009, background): “The technical details are beyond the scope of this paper, but accessible discussions be found in several places (e.g., Berk, 2008a: 222–231).”
Modeling heterogeneous treatment effects in large-scale experiments using Bayesian Additive Regression Trees (2010, methodology): “BART builds upon regression and classification tree models developed in the statistical learning literature (Morgan and Sonquist 1963; Breiman et al. 1984; Berk 2008; Izenman 2008; Hastie, Tibshirani, and Friedman 2009).”
Predictive Analytics in Information Systems Research (2011, unknown): “A popular method to obtain out-of-sample data is to initially partition the data randomly, using one part (the training set) to fit the empirical model, and the other (the holdout set) to assess the model’s predictive accuracy (Berk 2008, p.31; Hastie et al. 2008, p. 222).”
A survey of functional principal component analysis (2014, unknown): “61 - 63 for its proof ) .”
Small Area Estimation of the Homeless in Los Angeles: An Application of Cost-Sensitive stochastic Gradient Boosting (2010, unknown): “The stochastic gradient boosting algorithm in its most general form is provided below (Friedman, 2002; Ridgeway, 2007; Berk, 2008):6 1.”
A Primer on Propensity Score Analysis (2010, methodology): “algorithms like classification trees or ensemble methods.(9,10) No matter what method is used, all the available variables that were measured before treatment and are plausibly related to treatment and potential outcomes should be in the initial propensity score model.”
Forecasts of Violence to Inform Sentencing Decisions (2013, background): “But there is now a sound rationale and a wealth of results making conventional statistical models (e.g., logistic regression) less desirable for many applications, including some kinds of forecasting (Breiman 2001b; Hastie et al. 2009; Berk 2008b, 2012).”
Asymmetric Loss Functions for Forecasting in Criminal Justice Settings (2011, background): “No classifiers to date consistently classify and forecast more accurately than random forests (Breiman 2001; Berk 2008).”
Factors determining online sustainability reporting by local governments (2015, methodology): “This is suitable for our purposes and sufficient for the statistical analysis performed, for various reasons: first, it includes all the municipalities classified as major cities under Law 57/2003, on municipal transparency; second, the sample size used is similar to that applied in numerous special”
Modern Multivariate Statistical Techniques: Regression, Classification, and Manifold Learning (2008, background): “As with Berk (2008) (see also the review by Maindonald 2009), this book is broadly in the tradition of “statistical learning”.”
Random Forests for Global and Regional Crop Yield Predictions (2016, unknown): “Third, RF algorithms can use multiple types of predictors in a model more easily than traditional multiple linear or non-linear regressions can [20].”
A random forest based approach for predicting spreads in the primary catastrophe bond market (2020, methodology): “This default value was also suggested and used by many other authors, as Wang et al. (2018), Gr ¨ omping (2009), and Berk (2008) and therefore we also employ it as node size value here.”
When second best is good enough: a comparison between a true experiment and a regression discontinuity quasi-experiment (2010, methodology): “Random forests (Breiman, 2001; Berk 2008) was applied to a training data set of 30,000 observations.”