Online seminar series on statistical learning and related topics

The effects of CoViD-19 pervades through research communities across the globe, causing cancelled conferences, post-poned research visits, and suspended projects. Like many others, we have sought other opportunities for collaboration in spite of the current state of affairs and have therefore organized this online seminar series in statistical learning.

We are going to use zoom. The seminar will be 1 hour long, with 45 minutes allocated to the presentation itself and 15 minutes to discussions afterwards.

Here are some ground rules for these seminars:

- Please use your real name as a handle.
- Please mute your microphone whenever you are not speaking.
- If you wish to ask a question, use the
**raise hand**button in the**participants**window.

https://lu-se.zoom.us/j/65067339175

The seminar is held on a weekly basis on **fridays** and will run
at least until June 12. Each seminar starts at
16:30 CEST.

**Speaker**: Damian Brzyski**Title**: The adaptive incorporation of multiple sources of information in Brain Imaging via penalized optimization**Abstract**: The use of multiple sources of information in regression modeling has recently received a lot of attention in the statistical and brain imaging literature. This talk introduces a novel, fully-automatic statistical procedure that addresses the problem of linear regression coefficients estimation in the situation when the additional information about connectivities between variables is given. Our method, “Adaptive Information Merging Estimator for Regression” (AIMER) enables for the incorporation of multiple sources of such information as well as for the division of one source into pieces and determining their impact on the estimates. We performed extensive simulations to visualize the desired adjusting properties of our method and show its advantages over the existing approaches. We also applied AIMER to analyze structural brain imaging data and to reveal the association between cortical thickness and HIV-related outcomes.

**Speaker**: Aaron Molstad**Title**: Insights and algorithms for the multivariate square-root lasso**Abstract**: We study the multivariate square-root lasso, a method for fitting the multivariate response (i.e. multi-task) linear regression model with dependent errors. This estimator minimizes the nuclear norm of the residual matrix plus a convex penalty. Unlike some existing methods for multivariate response linear regression, which require explicit estimates of the error covariance matrix or its inverse, the multivariate square-root lasso criterion implicitly adapts to dependent errors and is convex. To justify the use of this estimator, we establish an error bound which illustrates that like the univariate square-root lasso, the multivariate square-root lasso is pivotal with respect to the unknown error covariance matrix. Based on our theory, we propose a simple tuning approach which requires fitting the model for only a single value of the tuning parameter, e.g., does not require cross-validation. We propose two algorithms to compute the estimator: a prox-linear alternating direction method of multipliers algorithm, and an accelerated first order algorithm which can be applied in certain cases. In both simulation studies and a genomic data application, we show that the multivariate square-root lasso can outperform more computationally intensive methods which estimate both the regression coefficient matrix and error precision matrix.

Date | Speaker | Title | Resources |
---|---|---|---|

May 29 | Wojchiech Rejchel | Fast and robust procedures in high-dimensional variable selection | presentation, slides, paper |

May 22 | Jaroslaw Harezlak | Brain Connectivity-Informed Adaptive Regularization for Generalized Outcomes | presentation, slides |

May 22 | Jaroslaw Harezlak | Wearable Devices - Statistical Learning to the Rescue | presentation, slides |

May 8 | Johan Larsson | The strong screening rule for SLOPE | presentation, slides, paper |

May 8 | Patrick Tardivel | Screening rules for the lasso | presentation |

Recordings of the talks on this seminar are hosted at https://vimeo.com/channels/statlearnseminar.

This seminar series is organized by The Department of Mathematics, Wrocław University and The Department of Statistics, Lund University.

- One World Probability Seminar
- One World Optimization Seminar
- One World ABC (Approximate Bayesian Computation) Seminar
- One World Mathematics of INformation, Data, and Signals (MINDS) Seminar
- One World Stochastic Numerics and Inverse Problems (SNIP) Seminar
- One World PDE Seminar
- One World Mathematical Game Theory Seminar
- One World Seminar: Mathematical Methods for Arbitrary Data Sources
- One World Cognitive Psychologie Seminar
- International Seminar on Selective Inference
- Online Causal Inference Seminar