# High-dimensional graphs and variable selection with the Lasso

Source: [High-dimensional graphs and variable selection with the Lasso](<https://www.sota2.com/research/paper/68fa530a1e209131ce980117>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T12:14:04.948Z

Authors: Nicolai Meinshausen, Peter B\\"uhlmann

Published: 2006-08-01T09:59:28.000Z

## Abstract

The pattern of zero entries in the inverse covariance matrix of a multivariate normal distribution corresponds to conditional independence restrictions between variables. Covariance selection aims at estimating those structural zeros from data. We show that neighborhood selection with the Lasso is a computationally attractive alternative to standard covariance selection for sparse high-dimensional graphs. Neighborhood selection estimates the conditional independence restrictions separately for each node in the graph and is hence equivalent to variable selection for Gaussian linear models. We show that the proposed neighborhood selection scheme is consistent for sparse high-dimensional graphs. Consistency hinges on the choice of the penalty parameter. The oracle value for optimal prediction does not lead to a consistent neighborhood estimate. Controlling instead the probability of falsely joining some distinct connectivity components of the graph, consistent estimation for sparse graphs is achieved (with exponential rates), even when the number of variables grows as the number of observations raised to an arbitrary power.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Variable Clustering on Simulation data](<https://www.sota2.com/research/sota/variable-clustering-on-simulation-data.md>): Variable Clustering; Simulation Data; Average AMI: 94.

- [Face Clustering on Extended Yale B](<https://www.sota2.com/research/sota/face-clustering-on-extended-yale-b.md>): Face Clustering; Extended Yale-B; Mean AMI: 0.41.

- [Subspace Clustering on Extended Yale B (3 standard splits + 20 random trials)](<https://www.sota2.com/research/sota/subspace-clustering-on-extended-yale-b-3-standard-splits-plus-20-random-trials.md>): Subspace Clustering; Extended Yale B (3 standard splits + 20 random trials); Mean AMI: 0.403.

- [Causal Structure Learning on Philadelphia Business Graph (Yelp)](<https://www.sota2.com/research/sota/causal-structure-learning-on-philadelphia-business-graph-yelp.md>): Causal Structure Learning; Philadelphia Business Graph (Yelp); SHD: 2.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/math/0608017>)

[PDF](<https://arxiv.org/pdf/math/0608017.pdf>)
