skip to main content

Search for: All records

Creators/Authors contains: "Zhao, T."

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Free, publicly-accessible full text available October 17, 2023
  2. Free, publicly-accessible full text available September 26, 2023
  3. The problem of continuous inverse optimal control (over finite time horizon) is to learn the unknown cost function over the sequence of continuous control variables from expert demonstrations. In this article, we study this fundamental problem in the framework of energy-based model, where the observed expert trajectories are assumed to be random samples from a probability density function defined as the exponential of the negative cost function up to a normalizing constant. The parameters of the cost function are learned by maximum likelihood via an “analysis by synthesis” scheme, which iterates (1) synthesis step: sample the synthesized trajectories from the current probability density using the Langevin dynamics via back-propagation through time, and (2) analysis step: update the model parameters based on the statistical difference between the synthesized trajectories and the observed trajectories. Given the fact that an efficient optimization algorithm is usually available for an optimal control problem, we also consider a convenient approximation of the above learning method, where we replace the sampling in the synthesis step by optimization. Moreover, to make the sampling or optimization more efficient, we propose to train the energy-based model simultaneously with a top-down trajectory generator via cooperative learning, where the trajectory generator ismore »used to fast initialize the synthesis step of the energy-based model. We demonstrate the proposed methods on autonomous driving tasks, and show that they can learn suitable cost functions for optimal control.« less
    Free, publicly-accessible full text available January 1, 2023
  4. Free, publicly-accessible full text available October 4, 2022
  5. Human trajectory prediction is critical for autonomous platforms like self-driving cars or social robots. We present a latent belief energy-based model (LB-EBM) for diverse human trajectory forecast. LB-EBM is a probabilistic model with cost function defined in the latent space to account for the movement history and social context. The low dimensionality of the latent space and the high expressivity of the EBM make it easy for the model to capture the multimodality of pedestrian trajectory distributions. LB-EBM is learned from expert demonstrations (i.e., human trajectories) projected into the latent space. Sampling from or optimizing the learned LB-EBM yields a belief vector which is used to make a path plan, which then in turn helps to predict a long-range trajectory. The effectiveness of LB-EBM and the two-step approach are supported by strong empirical results. Our model is able to make accurate, multi-modal, and social compliant trajectory predictions and improves over prior state-of-the-arts performance on the Stanford Drone trajectory prediction benchmark by 10:9% and on the ETH-UCY benchmark by 27:6%.
  6. We combine equation of state of dense matter up to twice nuclear saturation density (nsat = 0.16 fm−3 ) obtained using chiral effective field theory (χEFT), and recent observations of neutron stars to gain insights about the high-density matter encountered in their cores. A key element in our study is the recent Bayesian analysis of correlated EFT truncation errors based on order-byorder calculations up to next-to-next-to-next-to-leading order in the χEFT expansion. We refine the bounds on the maximum mass imposed by causality at high densities, and provide stringent limits on the maximum and minimum radii of ∼ 1.4 M and ∼ 2.0 M stars. Including χEFT predictions from nsat to 2 nsat reduces the permitted ranges of the radius of a 1.4 M star, R1.4, by ∼ 3.5 km. If observations indicate R1.4 < 11.2 km, our study implies that either the squared speed of sound c 2 s > 1/2 for densities above 2 nsat, or that χEFT breaks down below 2 nsat. We also comment on the nature of the secondary compact object in GW190814 with mass ' 2.6 M , and discuss the implications of massive neutron stars > 2.1 M (2.6 M ) in future radiomore »and gravitational-wave searches. Some form of strongly interacting matter with c 2 s > 0.35 (0.55) must be realized in the cores of such massive neutron stars. In the absence of phase transitions below 2 nsat, the small tidal deformability inferred from GW170817 lends support for the relatively small pressure predicted by χEFT for the baryon density nB in the range 1−2 nsat. Together they imply that the rapid stiffening required to support a high maximum mass should occur only when nB & 1.5 − 1.8 nsat.« less
  7. This paper proposes a new meta-learning method – named HARMLESS (HAwkes Relational Meta LEarning method for Short Sequences) for learning heterogeneous point process models from short event sequence data along with a relational network. Specifically, we propose a hierarchical Bayesian mixture Hawkes process model, which naturally incorporates the relational information among sequences into point process modeling. Compared with existing methods, our model can capture the underlying mixed-community patterns of the relational network, which simultaneously encourages knowledge sharing among sequences and facilitates adaptive learning for each individual sequence. We further propose an efficient stochastic variational meta expectation maximization algorithm that can scale to large problems. Numerical experiments on both synthetic and real data show that HARMLESS outperforms existing methods in terms of predicting the future events.