Michael Wieck-Sosa

Headshot

I am a third-year PhD Student in the Department of Statistics & Data Science at CMU. I am lucky to be advised by Aaditya Ramdas.

Here is my CV. You can reach me at mwiecksosa AT cmu DOT edu.

I work on nonstationary nonlinear time series analysis. I am inspired by many topics in statistics and machine learning, notably high-dimensional statistics, semiparametric statistics, nonparametric statistics, causal inference, online learning, and deep learning. I am especially interested in the connections between causality, forecasting, and control.

Working Papers

  1. Conditional independence testing with a single realization of a multivariate nonstationary nonlinear time series (with Michel F. C. Haddad and Aaditya Ramdas).
    • Summary: We introduce a general framework for conditional independence testing for nonstationary time series based on time-varying nonlinear regression and distribution-uniform strong Gaussian approximations.
    • Links: arXiv.
  2. Identifying relevant forecasting signals in unstable environments (with Michel F. C. Haddad and Aaditya Ramdas).
    • Summary: We develop a method for selecting variables for forecasting with nonstationary nonlinear time series.
  3. Simulation-based inference for nonlinear models of high-dimensional time series through random features (with Cosma Shalizi).
    • Summary: We propose a method for inferring the parameters of analytically intractable models for time series, which draws from nonstationary nonlinear time series theory.
  4. Deep learning for nonstationary nonlinear time series (with Wei Biao Wu).
    • Summary: We present a theoretical framework for estimating the time-varying regression functions of nonstationary nonlinear time series using deep neural networks.