Zhi Wang

Zhi Wang, Chunlin Chen, Daoyi Dong

While reinforcement learning (RL) algorithms are achieving state-of-the-art performance in various challenging tasks, they can easily encounter catastrophic forgetting or interference when faced with lifelong streaming information. In the paper, we propose a scalable lifelong RL method that dynamically expands the network capacity to accommodate new knowledge while preventing past memories from being perturbed. We use a Dirichlet process mixture to model the non-stationary task distribution, which captures task relatedness by estimating the likelihood of task-to-cluster assignments and clusters the task models in a latent space. We formulate the prior distribution of the mixture as a Chinese restaurant process (CRP) that instantiates new mixture components as needed. The update and expansion of the mixture are governed by the Bayesian non-parametric framework with an expectation maximization (EM) procedure, which dynamically adapts the model complexity without explicit task boundaries or heuristics. Moreover, we use the domain randomization technique to train robust prior parameters for the initialization of each task model in the mixture, thus the resulting model can better generalize and adapt to unseen tasks. With extensive experiments conducted on robot navigation and locomotion domains, we show that our method successfully facilitates scalable lifelong RL and outperforms relevant existing methods.

Donghan Xie, Zhi Wang, Chunlin Chen et al.

Multi-agent settings remain a fundamental challenge in the reinforcement learning (RL) domain due to the partial observability and the lack of accurate real-time interactions across agents. In this paper, we propose a new method based on local communication learning to tackle the multi-agent RL (MARL) challenge within a large number of agents coexisting. First, we design a new communication protocol that exploits the ability of depthwise convolution to efficiently extract local relations and learn local communication between neighboring agents. To facilitate multi-agent coordination, we explicitly learn the effect of joint actions by taking the policies of neighboring agents as inputs. Second, we introduce the mean-field approximation into our method to reduce the scale of agent interactions. To more effectively coordinate behaviors of neighboring agents, we enhance the mean-field approximation by a supervised policy rectification network (PRN) for rectifying real-time agent interactions and by a learnable compensation term for correcting the approximation bias. The proposed method enables efficient coordination as well as outperforms several baseline approaches on the adaptive traffic signal control (ATSC) task and the StarCraft II multi-agent challenge (SMAC).

Zichuan Liu, Yuanyang Zhu, Zhi Wang et al.

While achieving tremendous success in various fields, existing multi-agent reinforcement learning (MARL) with a black-box neural network makes decisions in an opaque manner that hinders humans from understanding the learned knowledge and how input observations influence decisions. In contrast, existing interpretable approaches usually suffer from weak expressivity and low performance. To bridge this gap, we propose MIXing Recurrent soft decision Trees (MIXRTs), a novel interpretable architecture that can represent explicit decision processes via the root-to-leaf path and reflect each agent's contribution to the team. Specifically, we construct a novel soft decision tree using a recurrent structure and demonstrate which features influence the decision-making process. Then, based on the value decomposition framework, we linearly assign credit to each agent by explicitly mixing individual action values to estimate the joint action value using only local observations, providing new insights into interpreting the cooperation mechanism. Theoretical analysis confirms that MIXRTs guarantee additivity and monotonicity in the factorization of joint action values. Evaluations on complex tasks like Spread and StarCraft II demonstrate that MIXRTs compete with existing methods while providing clear explanations, paving the way for interpretable and high-performing MARL systems.

Junyi Wang, Yuanyang Zhu, Zhi Wang et al.

Evolutionary reinforcement learning (ERL) algorithms recently raise attention in tackling complex reinforcement learning (RL) problems due to high parallelism, while they are prone to insufficient exploration or model collapse without carefully tuning hyperparameters (aka meta-parameters). In the paper, we propose a general meta ERL framework via bilevel optimization (BiERL) to jointly update hyperparameters in parallel to training the ERL model within a single agent, which relieves the need for prior domain knowledge or costly optimization procedure before model deployment. We design an elegant meta-level architecture that embeds the inner-level's evolving experience into an informative population representation and introduce a simple and feasible evaluation of the meta-level fitness function to facilitate learning efficiency. We perform extensive experiments in MuJoCo and Box2D tasks to verify that as a general framework, BiERL outperforms various baselines and consistently improves the learning performance for a diversity of ERL algorithms.

Shuzhao Xie, Yuan Xue, Yifei Zhu et al.

With the advancement of deep learning techniques, major cloud providers and niche machine learning service providers start to offer their cloud-based machine learning tools, also known as machine learning as a service (MLaaS), to the public. According to our measurement, for the same task, these MLaaSes from different providers have varying performance due to the proprietary datasets, models, etc. Federating different MLaaSes together allows us to improve the analytic performance further. However, naively aggregating results from different MLaaSes not only incurs significant momentary cost but also may lead to sub-optimal performance gain due to the introduction of possible false-positive results. In this paper, we propose Armol, a framework to federate the right selection of MLaaS providers to achieve the best possible analytic performance. We first design a word grouping algorithm to unify the output labels across different providers. We then present a deep combinatorial reinforcement learning based-approach to maximize the accuracy while minimizing the cost. The predictions from the selected providers are then aggregated together using carefully chosen ensemble strategies. The real-world trace-driven evaluation further demonstrates that Armol is able to achieve the same accuracy results with $67\%$ less inference cost.

Fanding Huang, Jingyan Jiang, Qinting Jiang et al.

Recent vision-language models (VLMs) face significant challenges in test-time adaptation to novel domains. While cache-based methods show promise by leveraging historical information, they struggle with both caching unreliable feature-label pairs and indiscriminately using single-class information during querying, significantly compromising adaptation accuracy. To address these limitations, we propose COSMIC (Clique-Oriented Semantic Multi-space Integration for CLIP), a robust test-time adaptation framework that enhances adaptability through multi-granular, cross-modal semantic caching and graph-based querying mechanisms. Our framework introduces two key innovations: Dual Semantics Graph (DSG) and Clique Guided Hyper-class (CGH). The Dual Semantics Graph constructs complementary semantic spaces by incorporating textual features, coarse-grained CLIP features, and fine-grained DINOv2 features to capture rich semantic relationships. Building upon these dual graphs, the Clique Guided Hyper-class component leverages structured class relationships to enhance prediction robustness through correlated class selection. Extensive experiments demonstrate COSMIC's superior performance across multiple benchmarks, achieving significant improvements over state-of-the-art methods: 15.81% gain on out-of-distribution tasks and 5.33% on cross-domain generation with CLIP RN-50. Code is available at github.com/hf618/COSMIC.

Thang Duong, Zhi Wang, Chicheng Zhang

We study lifelong learning in linear bandits, where a learner interacts with a sequence of linear bandit tasks whose parameters lie in an $m$-dimensional subspace of $\mathbb{R}^d$, thereby sharing a low-rank representation. Current literature typically assumes that the tasks are diverse, i.e., their parameters uniformly span the $m$-dimensional subspace. This assumption allows the low-rank representation to be learned before all tasks are revealed, which can be unrealistic in real-world applications. In this work, we present the first nontrivial result for sequential multi-task linear bandits without the task diversity assumption. We develop an algorithm that efficiently learns and transfers low-rank representations. When facing $N$ tasks, each played over $τ$ rounds, our algorithm achieves a regret guarantee of $\tilde{O}\big (Nm \sqrtτ + N^{\frac{2}{3}} τ^{\frac{2}{3}} d m^{\frac13} + Nd^2 + τm d \big)$ under the ellipsoid action set assumption. This result can significantly improve upon the baseline of $\tilde{O} \left (Nd \sqrtτ\right)$ that does not leverage the low-rank structure when the number of tasks $N$ is sufficiently large and $m \ll d$. We also demonstrate empirically on synthetic data that our algorithm outperforms baseline algorithms, which rely on the task diversity assumption.

Qingpeng Zhao, Yuanyang Zhu, Zichuan Liu et al.

In cooperative multi-agent reinforcement learning (MARL), the environmental stochasticity and uncertainties will increase exponentially when the number of agents increases, which puts hard pressure on how to come up with a compact latent representation from partial observation for boosting value decomposition. To tackle these issues, we propose a simple yet powerful method that alleviates partial observability and efficiently promotes coordination by introducing the UNit-wise attentive State Representation (UNSR). In UNSR, each agent learns a compact and disentangled unit-wise state representation outputted from transformer blocks, and produces its local action-value function. The proposed UNSR is used to boost the value decomposition with a multi-head attention mechanism for producing efficient credit assignment in the mixing network, providing an efficient reasoning path between the individual value function and joint value function. Experimental results demonstrate that our method achieves superior performance and data efficiency compared to solid baselines on the StarCraft II micromanagement challenge. Additional ablation experiments also help identify the key factors contributing to the performance of UNSR.

Chicheng Zhang, Zhi Wang

We study multi-task reinforcement learning (RL) in tabular episodic Markov decision processes (MDPs). We formulate a heterogeneous multi-player RL problem, in which a group of players concurrently face similar but not necessarily identical MDPs, with a goal of improving their collective performance through inter-player information sharing. We design and analyze an algorithm based on the idea of model transfer, and provide gap-dependent and gap-independent upper and lower bounds that characterize the intrinsic complexity of the problem.

Yuanyang Zhu, Zhi Wang, Chunlin Chen et al.

For real-world deployments, it is critical to allow robots to navigate in complex environments autonomously. Traditional methods usually maintain an internal map of the environment, and then design several simple rules, in conjunction with a localization and planning approach, to navigate through the internal map. These approaches often involve a variety of assumptions and prior knowledge. In contrast, recent reinforcement learning (RL) methods can provide a model-free, self-learning mechanism as the robot interacts with an initially unknown environment, but are expensive to deploy in real-world scenarios due to inefficient exploration. In this paper, we focus on efficient navigation with the RL technique and combine the advantages of these two kinds of methods into a rule-based RL (RuRL) algorithm for reducing the sample complexity and cost of time. First, we use the rule of wall-following to generate a closed-loop trajectory. Second, we employ a reduction rule to shrink the trajectory, which in turn effectively reduces the redundant exploration space. Besides, we give the detailed theoretical guarantee that the optimal navigation path is still in the reduced space. Third, in the reduced space, we utilize the Pledge rule to guide the exploration strategy for accelerating the RL process at the early stage. Experiments conducted on real robot navigation problems in hex-grid environments demonstrate that RuRL can achieve improved navigation performance.

Linghan Meng, Yanhui Li, Lin Chen et al.

The boom of DL technology leads to massive DL models built and shared, which facilitates the acquisition and reuse of DL models. For a given task, we encounter multiple DL models available with the same functionality, which are considered as candidates to achieve this task. Testers are expected to compare multiple DL models and select the more suitable ones w.r.t. the whole testing context. Due to the limitation of labeling effort, testers aim to select an efficient subset of samples to make an as precise rank estimation as possible for these models. To tackle this problem, we propose Sample Discrimination based Selection (SDS) to select efficient samples that could discriminate multiple models, i.e., the prediction behaviors (right/wrong) of these samples would be helpful to indicate the trend of model performance. To evaluate SDS, we conduct an extensive empirical study with three widely-used image datasets and 80 real world DL models. The experimental results show that, compared with state-of-the-art baseline methods, SDS is an effective and efficient sample selection method to rank multiple DL models.

Zhi Wang, Chunlin Chen, Daoyi Dong

Evolution strategies (ES), as a family of black-box optimization algorithms, recently emerge as a scalable alternative to reinforcement learning (RL) approaches such as Q-learning or policy gradient, and are much faster when many central processing units (CPUs) are available due to better parallelization. In this paper, we propose a systematic incremental learning method for ES in dynamic environments. The goal is to adjust previously learned policy to a new one incrementally whenever the environment changes. We incorporate an instance weighting mechanism with ES to facilitate its learning adaptation, while retaining scalability of ES. During parameter updating, higher weights are assigned to instances that contain more new knowledge, thus encouraging the search distribution to move towards new promising areas of parameter space. We propose two easy-to-implement metrics to calculate the weights: instance novelty and instance quality. Instance novelty measures an instance's difference from the previous optimum in the original environment, while instance quality corresponds to how well an instance performs in the new environment. The resulting algorithm, Instance Weighted Incremental Evolution Strategies (IW-IES), is verified to achieve significantly improved performance on challenging RL tasks ranging from robot navigation to locomotion. This paper thus introduces a family of scalable ES algorithms for RL domains that enables rapid learning adaptation to dynamic environments.

Zhi Wang, Xueying Tang, Jingchen Liu et al.

Response process data collected from human-computer interactive items contain rich information about respondents' behavioral patterns and cognitive processes. Their irregular formats as well as their large sizes make standard statistical tools difficult to apply. This paper develops a computationally efficient method for exploratory analysis of such process data. The new approach segments a lengthy individual process into a sequence of short subprocesses to achieve complexity reduction, easy clustering and meaningful interpretation. Each subprocess is considered a subtask. The segmentation is based on sequential action predictability using a parsimonious predictive model combined with the Shannon entropy. Simulation studies are conducted to assess performance of the new methods. We use the process data from PIAAC 2012 to demonstrate how exploratory analysis of process data can be done with the new approach.

Zhi Wang, Chunlin Chen, Daoyi Dong

A central capability of a long-lived reinforcement learning (RL) agent is to incrementally adapt its behavior as its environment changes, and to incrementally build upon previous experiences to facilitate future learning in real-world scenarios. In this paper, we propose LifeLong Incremental Reinforcement Learning (LLIRL), a new incremental algorithm for efficient lifelong adaptation to dynamic environments. We develop and maintain a library that contains an infinite mixture of parameterized environment models, which is equivalent to clustering environment parameters in a latent space. The prior distribution over the mixture is formulated as a Chinese restaurant process (CRP), which incrementally instantiates new environment models without any external information to signal environmental changes in advance. During lifelong learning, we employ the expectation maximization (EM) algorithm with online Bayesian inference to update the mixture in a fully incremental manner. In EM, the E-step involves estimating the posterior expectation of environment-to-cluster assignments, while the M-step updates the environment parameters for future learning. This method allows for all environment models to be adapted as necessary, with new models instantiated for environmental changes and old models retrieved when previously seen environments are encountered again. Experiments demonstrate that LLIRL outperforms relevant existing methods, and enables effective incremental adaptation to various dynamic environments for lifelong learning.

Xiao Ma, Liqin Zhao, Guan Huang et al.

Estimating post-click conversion rate (CVR) accurately is crucial for ranking systems in industrial applications such as recommendation and advertising. Conventional CVR modeling applies popular deep learning methods and achieves state-of-the-art performance. However it encounters several task-specific problems in practice, making CVR modeling challenging. For example, conventional CVR models are trained with samples of clicked impressions while utilized to make inference on the entire space with samples of all impressions. This causes a sample selection bias problem. Besides, there exists an extreme data sparsity problem, making the model fitting rather difficult. In this paper, we model CVR in a brand-new perspective by making good use of sequential pattern of user actions, i.e., impression -> click -> conversion. The proposed Entire Space Multi-task Model (ESMM) can eliminate the two problems simultaneously by i) modeling CVR directly over the entire space, ii) employing a feature representation transfer learning strategy. Experiments on dataset gathered from Taobao's recommender system demonstrate that ESMM significantly outperforms competitive methods. We also release a sampling version of this dataset to enable future research. To the best of our knowledge, this is the first public dataset which contains samples with sequential dependence of click and conversion labels for CVR modeling.