• Publications
  • Influence
Sample Efficient Policy Gradient Methods with Recursive Variance Reduction
TLDR
A novel policy gradient algorithm called SRVR-PG is proposed, which only requires one episode to find an $\epsilon$-approximate stationary point of the nonconcave performance function $J(\boldsymbol{\theta})$ and improves the existing result $O(1/\ep silon^{5/3})$ for stochastic variance reduced policy gradient algorithms by a factor of two.
An Improved Convergence Analysis of Stochastic Variance-Reduced Policy Gradient
TLDR
An improved convergence analysis of SVRPG is provided and it is shown that it can find an $\epsilon$-approximate stationary point of the performance function within $O(1/\ep silon^{5/3})$ trajectories, and sample complexity improves upon the best known result.
Global Convergence of Langevin Dynamics Based Algorithms for Nonconvex Optimization
TLDR
For the first time, it is proved that the global convergence guarantee for variance reduced stochastic gradient Langevin dynamics (VR-SGLD) to the almost minimizer after $\tilde O\big(\sqrt{n}d^5/(\lambda^4\epsilon^{5/2})\big)$ stoChastic gradient evaluations, which outperforms the gradient complexities of GLD and SGLD in a wide regime.
Stochastic Nested Variance Reduction for Nonconvex Optimization
TLDR
A new stochastic gradient descent algorithm based on nested variance reduction that improves the best known gradient complexity of SVRG and SCSG and achieves better gradient complexity than the state-of-the-art algorithms.
A Finite Time Analysis of Two Time-Scale Actor Critic Methods
TLDR
This work provides a non-asymptotic analysis for two time-scale actor-Critic methods under non-i.i.d. setting and proves that the actor-critic method is guaranteed to find a first-order stationary point.
Stochastic Nested Variance Reduced Gradient Descent for Nonconvex Optimization
TLDR
This work proposes a new stochastic gradient descent algorithm based on nested variance reduction that improves the best known gradient complexity of SVRG and the bestgradient complexity of SCSG.
Epidemic Model Guided Machine Learning for COVID-19 Forecasts in the United States
TLDR
The proposed SuEIR model is a variant of the SEIR model by taking into account the untested/unreported cases of COVID-19, and trained by machine learning algorithms based on the reported historical data to predict the peak date of active cases, and estimate the basic reproduction number.
Stochastic Variance-Reduced Cubic Regularized Newton Method
TLDR
This work shows that the proposed stochastic variance-reduced cubic regularized Newton method is guaranteed to converge to an approximately local minimum within $\tilde{O}(n^{4/5}/\epsilon^{3/2})$ second-order oracle calls, which outperforms the state-of-the-art cubic regularization algorithms including subsampled cubic regularizations.
Subsampled Stochastic Variance-Reduced Gradient Langevin Dynamics
TLDR
A nonasymptotic analysis of the convergence of SVRG-LD in 2-Wasserstein distance is provided, and it is shown that SVRg-LD enjoys a lower gradient complexity1 than SVRR-LD, when the sample size is large or the target accuracy requirement is moderate.
Evaluation of individual and ensemble probabilistic forecasts of COVID-19 mortality in the US
TLDR
This project systematically evaluates 23 models that regularly submitted forecasts of reported weekly incident COVID-19 mortality counts in the US at the state and national level and underscores the role that collaboration and active coordination between governmental public health agencies, academic modeling teams, and industry partners can play in developing modern modeling capabilities to support local, state, and federal response to outbreaks.
...
...