Least Squares Auto-Tuning google
Least squares is by far the simplest and most commonly applied computational method in many fields. In almost all applications, the least squares objective is rarely the true objective. We account for this discrepancy by parametrizing the least squares problem and automatically adjusting these parameters using an optimization algorithm. We apply our method, which we call least squares auto-tuning, to data fitting. …

GRU-ODE-Bayes google
Modeling real-world multidimensional time series can be particularly challenging when these are sporadically observed (i.e., sampling is irregular both in time and across dimensions)-such as in the case of clinical patient data. To address these challenges, we propose (1) a continuous-time version of the Gated Recurrent Unit, building upon the recent Neural Ordinary Differential Equations (Chen et al., 2018), and (2) a Bayesian update network that processes the sporadic observations. We bring these two ideas together in our GRU-ODE-Bayes method. We then demonstrate that the proposed method encodes a continuity prior for the latent process and that it can exactly represent the Fokker-Planck dynamics of complex processes driven by a multidimensional stochastic differential equation. Additionally, empirical evaluation shows that our method outperforms the state of the art on both synthetic data and real-world data with applications in healthcare and climate forecast. What is more, the continuity prior is shown to be well suited for low number of samples settings. …

Proportional Subdistribution Hazards (PSH) google
The proportional hazards model for the subdistribution that Fine and Gray (1999) propose aims at modeling the cumulative incidence of an event of interest.
“Proportional Hazards Model”


Covariate Balancing Propensity Score (CBPS) google
Implements the covariate balancing propensity score (CBPS) proposed by Imai and Ratkovic (2014) <DOI:10.1111/rssb.12027>. The propensity score is estimated such that it maximizes the resulting covariate balance as well as the prediction of treatment assignment. The method, therefore, avoids an iteration between model fitting and balance checking. The package also implements several extensions of the CBPS beyond the cross-sectional, binary treatment setting. The current version implements the CBPS for longitudinal settings so that it can be used in conjunction with marginal structural models from Imai and Ratkovic (2015) <DOI:10.1080/01621459.2014.956872>, treatments with three- and four- valued treatment variables, continuous-valued treatments from Fong, Hazlett, and Imai (2015) <http://…/CBGPS.pdf>, and the situation with multiple distinct binary treatments administered simultaneously. In the future it will be extended to other settings including the generalization of experimental and instrumental variable estimates. Recently we have added the optimal CBPS which chooses the optimal balancing function and results in doubly robust and efficient estimator for the treatment effect as well as high dimensional CBPS when a large number of covariates exist. …