Nicolai Meinshausen

Nicolai Meinshausen

9 packages on CRAN

howmany

cran
99.99th

Percentile

When testing multiple hypotheses simultaneously, this package provides functionality to calculate a lower bound for the number of correct rejections (as a function of the number of rejected hypotheses), which holds simultaneously -with high probability- for all possible number of rejections. As a special case, a lower bound for the total number of false null hypotheses can be inferred. Dependent test statistics can be handled for multiple tests of associations. For independent test statistics, it is sufficient to provide a list of p-values.

99.99th

Percentile

Confidence intervals for causal effects, using data collected in different experimental or environmental conditions. Hidden variables can be included in the model with a more experimental version.

99.99th

Percentile

Node harvest is a simple interpretable tree-like estimator for high-dimensional regression and classification. A few nodes are selected from an initially large ensemble of nodes, each associated with a positive weight. New observations can fall into one or several nodes and predictions are the weighted average response across all these groups. The package offers visualization of the estimator. Predictions can return the nodes a new observation fell into, along with the mean response of training observations in each node, offering a simple explanation of the prediction.

99.99th

Percentile

Low-dimensional embedding, using Random Forests for multiclass classification

relaxo

cran
99.99th

Percentile

Relaxed Lasso is a generalisation of the Lasso shrinkage technique for linear regression. Both variable selection and parameter estimation is achieved by regular Lasso, yet both steps do not necessarily use the same penalty parameter. The results include all standard Lasso solutions but allow often for sparser models while having similar or even slightly better predictive performance if many predictor variables are present. The package depends on the LARS package.

bastah

cran
99.99th

Percentile

Big data statistical analysis for high-dimensional models is made possible by modifying lasso.proj() in 'hdi' package by replacing its nodewise-regression with sparse precision matrix computation using 'BigQUIC'.

99.99th

Percentile

Unified interface for the estimation of causal networks, including the methods 'backShift' (from package 'backShift'), 'bivariateANM' (bivariate additive noise model), 'bivariateCAM' (bivariate causal additive model), 'CAM' (causal additive model) (from package 'CAM'; the package is temporarily unavailable on the CRAN repository; formerly available versions can be obtained from the archive), 'hiddenICP' (invariant causal prediction with hidden variables), 'ICP' (invariant causal prediction) (from package 'InvariantCausalPrediction'), 'GES' (greedy equivalence search), 'GIES' (greedy interventional equivalence search), 'LINGAM', 'PC' (PC Algorithm), 'FCI' (fast causal inference), 'RFCI' (really fast causal inference) (all from package 'pcalg') and regression.

hdi

cran
99.99th

Percentile

Implementation of multiple approaches to perform inference in high-dimensional models.

99.99th

Percentile

Quantile Regression Forests is a tree-based ensemble method for estimation of conditional quantiles. It is particularly well suited for high-dimensional data. Predictor variables of mixed classes can be handled. The package is dependent on the package 'randomForest', written by Andy Liaw.