Designing learning agents that explore efficiently in a complex environment has been widely recognized as a fundamental challenge in reinforcement learning. While a number of works have demonstrated the effectiveness of techniques based on randomized value functions on a single agent, it remains unclear, from a theoretical point of view, whether injecting randomization can help a society of agents concurently explore an environment. The theoretical results established in this work tender an affirmative answer to this question. We adapt the concurrent learning framework to randomized least-squares value iteration (RLSVI) with aggregated state representation. We demonstrate polynomial worst-case regret bounds in both finite- and infinite-horizon environments. In both setups the per-agent regret decreases at an optimal rate of $\Theta\left(\frac{1}{\sqrt{N}}\right)$, highlighting the advantage of concurent learning. Our algorithm exhibits significantly lower space complexity compared to Russo (2019) and Agrawal et. al (2021). We reduce the space complexity by a factor of $K$ while incurring only a $\sqrt{K}$ increase in the worst-case regret bound, compared to Russo (2019) and Agrawal et. al (2021). Interestingly, our algorithm improves the worst-case regret bound of Russo (2019) by a factor of $H^{1/2}$, matching the improvement in Agrawal et. al (2021). However, this result is achieved through a fundamentally different algorithmic enhancement and proof technique. Additionally, we conduct numerical experiments to demonstrate our theoretical findings.

Particle-based Bayesian deep learning often requires a similarity metric to compare two networks. However, naive similarity metrics lack permutation invariance and are inappropriate for comparing networks. Centered Kernel Alignment (CKA) on feature kernels has been proposed to compare deep networks but has not been used as an optimization objective in Bayesian deep learning. In this paper, we explore the use of CKA in Bayesian deep learning to generate diverse ensembles and hypernetworks that output a network posterior. Noting that CKA projects kernels onto a unit hypersphere and that directly optimizing the CKA objective leads to diminishing gradients when two networks are very similar. We propose adopting the approach of hyperspherical energy (HE) on top of CKA kernels to address this drawback and improve training stability. Additionally, by leveraging CKA-based feature kernels, we derive feature repulsive terms applied to synthetically generated outlier examples. Experiments on both diverse ensembles and hypernetworks show that our approach significantly outperforms baselines in terms of uncertainty quantification in both synthetic and realistic outlier detection tasks.

Behavior constrained policy optimization has been demonstrated to be a successful paradigm for tackling Offline Reinforcement Learning. By exploiting historical transitions, a policy is trained to maximize a learned value function while constrained by the behavior policy to avoid a significant distributional shift. In this paper, we propose our closed-form policy improvement operators. We make a novel observation that the behavior constraint naturally motivates the use of first-order Taylor approximation, leading to a linear approximation of the policy objective. Additionally, as practical datasets are usually collected by heterogeneous policies, we model the behavior policies as a Gaussian Mixture and overcome the induced optimization difficulties by leveraging the LogSumExp’s lower bound and Jensen’s Inequality, giving rise to a closed-form policy improvement operator. We instantiate both one-step and iterative offline RL algorithms with our novel policy improvement operators and empirically demonstrate their effectiveness over state-of-the-art algorithms on the standard D4RL benchmark. Our code is available at https://cfpi-icml23.github.io/.

Reinforcement learning (RL) has demonstrated remarkable achievements in simulated environments. However, carrying this success to real environments requires the important attribute of robustness, which the existing RL algorithms often lack as they assume that the future deployment environment is the same as the training environment (i.e. simulator) in which the policy is learned. This assumption often does not hold due to the discrepancy between the simulator and the real environment and, as a result, and hence renders the learned policy fragile when deployed. In this paper, we propose a novel distributionally robust $Q$-learning algorithm that learns the best policy in the worst distributional perturbation of the environment. Our algorithm first transforms the infinite-dimensional learning problem (since the environment MDP perturbation lies in an infinite-dimensional space) into a finite-dimensional dual problem and subsequently uses a multi-level Monte-Carlo scheme to approximate the dual value using samples from the simulator. Despite the complexity, we show that the resulting distributionally robust $Q$-learning algorithm asymptotically converges to optimal worst-case policy, thus making it robust to future environment changes. Simulation results further demonstrate its strong empirical robustness.

We consider the concurrent reinforcement learning problem where $n$ agents simultaneously learn to make decisions in the same environment by sharing experience with each other. Existing works in this emerging area have empirically demonstrated that Thompson sampling (TS) based algorithms provide a particularly attractive alternative for inducing cooperation, because each agent can independently sample a belief environment (and compute a corresponding optimal policy) from the joint posterior computed by aggregating all agents' data , which induces diversity in exploration among agents while benefiting shared experience from all agents. However, theoretical guarantees in this area remain under-explored; in particular, no regret bound is known on TS based concurrent RL algorithms. In this paper, we fill in this gap by considering two settings. In the first, we study the simple finite-horizon episodic RL setting, where TS is naturally adapted into the concurrent setup by having each agent sample from the current joint posterior at the beginning of each episode. We establish a $\tilde{O}(HS\sqrt{\frac{AT}{n}})$ per-agent regret bound, where $H$ is the horizon of the episode, $S$ is the number of states, $A$ is the number of actions, $T$ is the number of episodes and $n$ is the number of agents. In the second setting, we consider the infinite-horizon RL problem, where a policy is measured by its long-run average reward. Here, despite not having natural episodic breakpoints, we show that by a doubling-horizon schedule, we can adapt TS to the infinite-horizon concurrent learning setting to achieve a regret bound of $\tilde{O}(DS\sqrt{ATn})$, where $D$ is the standard notion of diameter of the underlying MDP and $T$ is the number of timesteps. Note that in both settings, the per-agent regret decreases at an optimal rate of $\Theta(\frac{1}{\sqrt{n}})$, which manifests the power of cooperation in concurrent RL.

While reinforcement learning has witnessed tremendous success recently in a wide range of domains, robustness–or the lack thereof–remains an important issue that remains inadequately addressed. In this paper, we provide a distributionally robust formulation of offline learning policy in tabular RL that aims to learn a policy from historical data (collected by some other behavior policy) that is robust to the future environment arising as a perturbation of the training environment. We first develop a novel policy evaluation scheme that accurately estimates the robust value (i.e. how robust it is in a perturbed environment) of any given policy and establish its finite-sample estimation error. Building on this, we then develop a novel and minimax-optimal distributionally robust learning algorithm that achieves $O_P\left(1/\sqrt{n}\right)$ regret, meaning that with high probability, the policy learned from using $n$ training data points will be $O\left(1/\sqrt{n}\right)$ close to the optimal distributionally robust policy. Finally, our simulation results demonstrate the superiority of our distributionally robust approach compared to non-robust RL algorithms.

We propose a novel Siamese Natural Language Tracker (SNLT), which brings the advancements in visual tracking to the tracking by natural language (NL) descriptions task. The proposed SNLT is applicable to a wide range of Siamese trackers, providing a new class of baselines for the tracking by NL task and promising future improvements from the advancements of Siamese trackers. The carefully designed architecture of the Siamese Natural Language Region Proposal Network (SNL-RPN), together with the Dynamic Aggregation of vision and language modalities, is introduced to perform the tracking by NL task. Empirical results over tracking benchmarks with NL annotations show that the proposed SNLT improves Siamese trackers by 3 to 7 percentage points with a slight tradeoff of speed. The proposed SNLT outperforms all NL trackers to-date and is competitive among state-of-the-art real-time trackers on LaSOT benchmarks while running at 50 frames per second on a single GPU. Code for this work is available at https://github.com/fredfung007/snlt.

Recently, particle-based variational inference (ParVI) methods have gained interest because they can avoid arbitrary parametric assumptions that are common in variational inference. However, many ParVI approaches do not allow arbitrary sampling from the posterior, and the few that do allow such sampling suffer from suboptimality. This work proposes a new method for learning to approximately sample from the posterior distribution. We construct a neural sampler that is trained with the functional gradient of the KL-divergence between the empirical sampling distribution and the target distribution, assuming the gradient resides within a reproducing kernel Hilbert space. Our generative ParVI (GPVI) approach maintains the asymptotic performance of ParVI methods while offering the flexibility of a generative sampler. Through carefully constructed experiments, we show that GPVI outperforms previous generative ParVI methods such as amortized SVGD, and is competitive with ParVI as well as gold-standard approaches like Hamiltonian Monte Carlo for fitting both exactly known and intractable target distributions.

Efficient exploration remains a challenging problem in reinforcement learning, especially for those tasks where rewards from environments are sparse. In this work, we introduce an exploration approach based on a novel implicit generative modeling algorithm to estimate a Bayesian uncertainty of the agent’s belief of the environment dynamics. Each random draw from our generative model is a neural network that instantiates the dynamic function, hence multiple draws would approximate the posterior, and the variance in the predictions based on this posterior is used as an intrinsic reward for exploration. We design a training algorithm for our generative model based on the amortized Stein Variational Gradient Descent. In experiments, we demonstrate the effectiveness of this exploration algorithm in both pure exploration tasks and a downstream task, comparing with state-of-the-art intrinsic reward-based exploration approaches, including two recent approaches based on an ensemble of dynamic models. In challenging exploration tasks, our implicit generative model consistently outperforms competing approaches regarding data efficiency in exploration.

In this work, we argue that conditioning on the natural language (NL) description of a target provides information for longer-term invariance, and thus helps cope with typical tracking challenges. However, deriving a formulation to combine the strengths of appearance-based tracking with the language modality is not straightforward. Therefore, we propose a novel deep tracking-by-detection formulation that can take advantage of NL descriptions. Regions that are related to the given NL description are generated by a proposal network during the detection stage of the tracker. Our LSTM based tracker then predicts the update of the target from regions proposed by the NL based detection stage. Our method runs at over 30 fps on a single GPU. In benchmarks, our method is competitive with state of the art trackers that employ bounding boxes for initialization, while it outperforms all other trackers on targets given unambiguous and precise language annotations. When conditioned on NL descriptions only, our model doubles the performance of the previous best attempt [25].

Regularization plays a crucial role in supervised learning. Most existing methods enforce a global regularization in a structure agnostic manner. In this paper, we initiate a new direction and propose to enforce the structural simplicity of the classification boundary by regularizing over its topological complexity. In particular, our measurement of topological complexity incorporates the importance of topological features (e.g., connected components, handles, and so on) in a meaningful manner, and provides a direct control over spurious topological structures. We incorporate the new measurement as a topological penalty in training classifiers. We also propose an efficient algorithm to compute the gradient of such penalty. Our method provides a novel way to topologically simplify the global structure of the model, without having to sacrifice too much of the flexibility of the model. We demonstrate the effectiveness of our new topological regularizer on a range of synthetic and real-world datasets.

Conventional unsupervised domain adaptation (UDA) assumes that training data are sampled from a single domain. This neglects the more practical scenario where training data are collected from multiple sources, requiring multi-source domain adaptation. We make three major contributions towards addressing this problem. First, we collect and annotate by far the largest UDA dataset, called DomainNet, which contains six domains and about 0.6 million images distributed among 345 categories, addressing the gap in data availability for multi-source UDA research. Second, we propose a new deep learning approach, Moment Matching for Multi-Source Domain Adaptation (M3SDA), which aims to transfer knowledge learned from multiple labeled source domains to an unlabeled target domain by dynamically aligning moments of their feature distributions. Third, we provide new theoretical insights specifically for moment matching approaches in both single and multiple source domain adaptation. Extensive experiments are conducted to demonstrate the power of our new dataset in benchmarking state-of-the-art multi-source domain adaptation methods, as well as the advantage of our proposed model. Dataset and Code are available at http://ai.bu.edu/M3SDA/.

We study the problem of supervised learning for both binary and multiclass classification from a unified geometric perspective. In particular, we propose a geometric regularization technique to find the submanifold corresponding to an estimator of the class probability P(y|\vec x). The regularization term measures the volume of this submanifold, based on the intuition that overfitting produces rapid local oscillations and hence large volume of the estimator. This technique can be applied to regularize any classification function that satisfies two requirements: firstly, an estimator of the class probability can be obtained; secondly, first and second derivatives of the class probability estimator can be calculated. In experiments, we apply our regularization technique to standard loss functions for classification, our RBF-based implementation compares favorably to widely used regularization methods for both binary and multiclass classification.

We propose a Bayesian framework for recursively estimating the classifier weights in online learning of a classifier ensemble. In contrast with past methods, such as stochastic gradient descent or online boosting, our framework estimates the weights in terms of evolving posterior distributions. For a specified class of loss functions, we show that it is possible to formulate a suitably defined likelihood function and hence use the posterior distribution as an approximation to the global empirical loss minimizer. If the stream of training data is sampled from a stationary process, we can also show that our framework admits a superior rate of convergence to the expected loss minimizer than is possible with standard stochastic gradient descent. In experiments with real-world datasets, our formulation often performs better than online boosting algorithms.

We propose a randomized ensemble algorithm to model the time-varying appearance of an object for visual tracking. In contrast with previous online methods for updating classifier ensembles in tracking-by-detection, the weight vector that combines weak classifiers is treated as a random variable and the posterior distribution for the weight vector is estimated in a Bayesian manner. In essence, the weight vector is treated as a distribution that reflects the confidence among the weak classifiers used to construct and adapt the classifier ensemble. The resulting formulation models the time-varying discriminative ability among weak classifiers so that the ensembled strong classifier can adapt to the varying appearance, backgrounds, and occlusions. The formulation is tested in a tracking-by-detection implementation. Experiments on 28 challenging benchmark videos demonstrate that the proposed method can achieve results comparable to and often better than those of state-of-the-art approaches.

There is plenty of structured information (such as lines and planes) in urban scenes. Considering this, we propose a new method for making use of this information to enhance the reconstruction of urban outdoor scenes. Structured information (collinearity and coplanarity) is extracted from images by performing line detection and color image segmentation, which is used as hypothetic constraints of the 3D structure. In refining stage, we first build PCA subspaces for each structured components (collinear and coplanar point sets), during which the former hypothetic structure information is further inspected by the initial 3D structure. Then we iteratively update the structure through EM estimation. Experiments show that this method effectively improves the accuracy and robustness of reconstruction of urban scenes.