Search

Normalization layers have recently experienced a renaissance in the deep reinforcement learning and continual learning literature, with several works highlighting diverse benefits such as improving loss landscape conditioning and combatting overestimation bias. However, normalization brings with it a subtle but important side effect: an equivalence between growth in the norm of the network parameters and decay in the effective learning rate. This becomes problematic in continual learning settings, where the resulting effective learning rate schedule may decay to near zero too quickly relative to the timescale of the learning problem. We propose to make the learning rate schedule explicit with a simple re-parameterization which we call Normalize-and-Project (NaP), which couples the insertion of normalization layers with weight projection, ensuring that the effective learning rate remains constant throughout training. This technique reveals itself as a powerful analytical tool to better understand learning rate schedules in deep reinforcement learning, and as a means of improving robustness to nonstationarity in synthetic plasticity loss benchmarks along with both the single-task and sequential variants of the Arcade Learning Environment. We also show that our approach can be easily applied to popular architectures such as ResNets and transformers while recovering and in some cases even slightly improving the performance of the base model in common stationary benchmarks.

In healthcare applications, there is a growing need to develop machine learning models that use data from a single source, such as that from a wrist wearable device, to monitor physical activities, assess health risks, and provide immediate health recommendations or interventions. However, the limitation of using single-source data often compromises the model's accuracy, as it fails to capture the full scope of human activities. While a more comprehensive dataset can be gathered in a lab setting using multiple sensors attached to various body parts, this approach is not practical for everyday use due to the impracticality of wearing multiple sensors. To address this challenge, we introduce a transfer learning framework that optimizes machine learning models for everyday applications by leveraging multi-source data collected in a laboratory setting. We introduce a novel metric to leverage the inherent relationship between these multiple data sources, as they are all paired to capture aspects of the same physical activity. Through numerical experiments, our framework outperforms existing methods in classification accuracy and robustness to noise, offering a promising avenue for the enhancement of daily activity monitoring.

Reinforcement learning from human feedback (RLHF) is the canonical framework for large language model alignment. However, rising popularity in offline alignment algorithms challenge the need for on-policy sampling in RLHF. Within the context of reward over-optimization, we start with an opening set of experiments that demonstrate the clear advantage of online methods over offline methods. This prompts us to investigate the causes to the performance discrepancy through a series of carefully designed experimental ablations. We show empirically that hypotheses such as offline data coverage and data quality by itself cannot convincingly explain the performance difference. We also find that while offline algorithms train policy to become good at pairwise classification, it is worse at generations; in the meantime the policies trained by online algorithms are good at generations while worse at pairwise classification. This hints at a unique interplay between discriminative and generative capabilities, which is greatly impacted by the sampling process. Lastly, we observe that the performance discrepancy persists for both contrastive and non-contrastive loss functions, and appears not to be addressed by simply scaling up policy networks. Taken together, our study sheds light on the pivotal role of on-policy sampling in AI alignment, and hints at certain fundamental challenges of offline alignment algorithms.

Pre-trained large language models (LLM) have emerged as a powerful tool for simulating various scenarios and generating output given specific instructions and multimodal input. In this work, we analyze the specific use of LLM to enhance a classical supervised machine learning method for classification problems. We propose a few approaches to integrate LLM into a classical machine learning estimator to further enhance the prediction performance. We examine the performance of the proposed approaches through both standard supervised learning binary classification tasks, and a transfer learning task where the test data observe distribution changes compared to the training data. Numerical experiments using four publicly available datasets are conducted and suggest that using LLM to enhance classical machine learning estimators can provide significant improvement on prediction performance.

Drug discovery is a complex process that involves sequentially screening and examining a vast array of molecules to identify those with the target properties. This process, also referred to as sequential experimentation, faces challenges due to the vast search space, the rarity of target molecules, and constraints imposed by limited data and experimental budgets. To address these challenges, we introduce a human-in-the-loop framework for sequential experiments in drug discovery. This collaborative approach combines human expert knowledge with deep learning algorithms, enhancing the discovery of target molecules within a specified experimental budget. The proposed algorithm processes experimental data to recommend both promising molecules and those that could improve its performance to human experts. Human experts retain the final decision-making authority based on these recommendations and their domain expertise, including the ability to override algorithmic recommendations. We applied our method to drug discovery tasks using real-world data and found that it consistently outperforms all baseline methods, including those which rely solely on human or algorithmic input. This demonstrates the complementarity between human experts and the algorithm. Our results provide key insights into the levels of humans' domain knowledge, the importance of meta-knowledge, and effective work delegation strategies. Our findings suggest that such a framework can significantly accelerate the development of new vaccines and drugs by leveraging the best of both human and artificial intelligence.

We provide a preliminary study on utilizing GPU (Graphics Processing Unit) to accelerate computation for three simulation optimization tasks with either first-order or second-order algorithms. Compared to the implementation using only CPU (Central Processing Unit), the GPU implementation benefits from computational advantages of parallel processing for large-scale matrices and vectors operations. Numerical experiments demonstrate computational advantages of utilizing GPU implementation in simulation optimization problems, and show that such advantage comparatively further increase as the problem scale increases.

With the advancement in generative language models, the selection of prompts has gained significant attention in recent years. A prompt is an instruction or description provided by the user, serving as a guide for the generative language model in content generation. Despite existing methods for prompt selection that are based on human labor, we consider facilitating this selection through simulation optimization, aiming to maximize a pre-defined score for the selected prompt. Specifically, we propose a two-stage framework. In the first stage, we determine a feasible set of prompts in sufficient numbers, where each prompt is represented by a moderate-dimensional vector. In the subsequent stage for evaluation and selection, we construct a surrogate model of the score regarding the moderate-dimensional vectors that represent the prompts. We propose sequentially selecting the prompt for evaluation based on this constructed surrogate model. We prove the consistency of the sequential evaluation procedure in our framework. We also conduct numerical experiments to demonstrate the efficacy of our proposed framework, providing practical instructions for implementation.

Ensuring alignment of language models' outputs with human preferences is critical to guarantee a useful, safe, and pleasant user experience. Thus, human alignment has been extensively studied recently and several methods such as Reinforcement Learning from Human Feedback (RLHF), Direct Policy Optimisation (DPO) and Sequence Likelihood Calibration (SLiC) have emerged. In this paper, our contribution is two-fold. First, we show the equivalence between two recent alignment methods, namely Identity Policy Optimisation (IPO) and Nash Mirror Descent (Nash-MD). Second, we introduce a generalisation of IPO, named IPO-MD, that leverages the regularised sampling approach proposed by Nash-MD. This equivalence may seem surprising at first sight, since IPO is an offline method whereas Nash-MD is an online method using a preference model. However, this equivalence can be proven when we consider the online version of IPO, that is when both generations are sampled by the online policy and annotated by a trained preference model. Optimising the IPO loss with such a stream of data becomes then equivalent to finding the Nash equilibrium of the preference model through self-play. Building on this equivalence, we introduce the IPO-MD algorithm that generates data with a mixture policy (between the online and reference policy) similarly as the general Nash-MD algorithm. We compare online-IPO and IPO-MD to different online versions of existing losses on preference data such as DPO and SLiC on a summarisation task.

Underpinning the past decades of work on the design, initialization, and optimization of neural networks is a seemingly innocuous assumption: that the network is trained on a \textit{stationary} data distribution. In settings where this assumption is violated, e.g.\ deep reinforcement learning, learning algorithms become unstable and brittle with respect to hyperparameters and even random seeds. One factor driving this instability is the loss of plasticity, meaning that updating the network's predictions in response to new information becomes more difficult as training progresses. While many recent works provide analyses and partial solutions to this phenomenon, a fundamental question remains unanswered: to what extent do known mechanisms of plasticity loss overlap, and how can mitigation strategies be combined to best maintain the trainability of a network? This paper addresses these questions, showing that loss of plasticity can be decomposed into multiple independent mechanisms and that, while intervening on any single mechanism is insufficient to avoid the loss of plasticity in all cases, intervening on multiple mechanisms in conjunction results in highly robust learning algorithms. We show that a combination of layer normalization and weight decay is highly effective at maintaining plasticity in a variety of synthetic nonstationary learning tasks, and further demonstrate its effectiveness on naturally arising nonstationarities, including reinforcement learning in the Arcade Learning Environment.

Offline preference optimization allows fine-tuning large models directly from offline data, and has proved effective in recent alignment practices. We propose generalized preference optimization (GPO), a family of offline losses parameterized by a general class of convex functions. GPO enables a unified view over preference optimization, encompassing existing algorithms such as DPO, IPO and SLiC as special cases, while naturally introducing new variants. The GPO framework also sheds light on how offline algorithms enforce regularization, through the design of the convex function that defines the loss. Our analysis and experiments reveal the connections and subtle differences between the offline regularization and the KL divergence regularization intended by the canonical RLHF formulation. In a controlled setting akin to Gao et al 2023, we also show that different GPO variants achieve similar trade-offs between regularization and performance, though the optimal values of hyper-parameter might differ as predicted by theory. In all, our results present new algorithmic toolkits and empirical insights to alignment practitioners., Comment: Accepted at ICML 2023 main conference

This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. We discuss our approach toward post-training and deploying Gemini models responsibly to users through services including Gemini, Gemini Advanced, Google AI Studio, and Cloud Vertex AI.

Utilizing covariate information has been a powerful approach to improve the efficiency and accuracy for causal inference, which support massive amount of randomized experiments run on data-driven enterprises. However, state-of-art approaches can become practically unreliable when the dimension of covariate increases to just 50, whereas experiments on large platforms can observe even higher dimension of covariate. We propose a machine-learning-assisted covariate representation approach that can effectively make use of historical experiment or observational data that are run on the same platform to understand which lower dimensions can effectively represent the higher-dimensional covariate. We then propose design and estimation methods with the covariate representation. We prove statistically reliability and performance guarantees for the proposed methods. The empirical performance is demonstrated using numerical experiments.

The $\textit{Abelian Sandpile}$ model is a well-known model used in exploring $\textit{self-organized criticality}$. Despite a large amount of work on other aspects of sandpiles, there have been limited results in efficiently computing the terminal state, known as the $\textit{sandpile prediction}$ problem. On graphs with special structures, we present algorithms that compute the terminal configurations for sandpile instances in $O(n \log n)$ time on trees and $O(n)$ time on paths, where $n$ is the number of vertices. Our algorithms improve the previous best runtime of $O(n \log^5 n)$ on trees [Ramachandran-Schild SODA '17] and $O(n \log n)$ on paths [Moore-Nilsson '99]. To do so, we move beyond the simulation of individual events by directly computing the number of firings for each vertex. The computation is accelerated using splittable binary search trees. In addition, we give algorithms in $O(n)$ time on cliques and $O(n \log^2 n)$ time on pseudotrees. On general graphs, we propose a fast algorithm under the setting where the number of chips $N$ could be arbitrarily large. We obtain a $\log N$ dependency, improving over the $\mathtt{poly}(N)$ dependency in purely simulation-based algorithms. Our algorithm also achieves faster performance on various types of graphs, including regular graphs, expander graphs, and hypercubes. We also provide a reduction that enables us to decompose the input sandpile into several smaller instances and solve them separately., Comment: 68 pages, submitted to FOCS24

We study the trade-off between expectation and tail risk for regret distribution in the stochastic multi-armed bandit problem. We fully characterize the interplay among three desired properties for policy design: worst-case optimality, instance-dependent consistency, and light-tailed risk. We show how the order of expected regret exactly affects the decaying rate of the regret tail probability for both the worst-case and instance-dependent scenario. A novel policy is proposed to characterize the optimal regret tail probability for any regret threshold. Concretely, for any given $\alpha\in[1/2, 1)$ and $\beta\in[0, \alpha]$, our policy achieves a worst-case expected regret of $\tilde O(T^\alpha)$ (we call it $\alpha$-optimal) and an instance-dependent expected regret of $\tilde O(T^\beta)$ (we call it $\beta$-consistent), while enjoys a probability of incurring an $\tilde O(T^\delta)$ regret ($\delta\geq\alpha$ in the worst-case scenario and $\delta\geq\beta$ in the instance-dependent scenario) that decays exponentially with a polynomial $T$ term. Such decaying rate is proved to be best achievable. Moreover, we discover an intrinsic gap of the optimal tail rate under the instance-dependent scenario between whether the time horizon $T$ is known a priori or not. Interestingly, when it comes to the worst-case scenario, this gap disappears. Finally, we extend our proposed policy design to (1) a stochastic multi-armed bandit setting with non-stationary baseline rewards, and (2) a stochastic linear bandit setting. Our results reveal insights on the trade-off between regret expectation and regret tail risk for both worst-case and instance-dependent scenarios, indicating that more sub-optimality and inconsistency leave space for more light-tailed risk of incurring a large regret, and that knowing the planning horizon in advance can make a difference on alleviating tail risks., Comment: arXiv admin note: text overlap with arXiv:2206.02969

We formulate, analyze and solve the problem of best arm identification with fairness constraints on subpopulations (BAICS). Standard best arm identification problems aim at selecting an arm that has the largest expected reward where the expectation is taken over the entire population. The BAICS problem requires that an selected arm must be fair to all subpopulations (e.g., different ethnic groups, age groups, or customer types) by satisfying constraints that the expected reward conditional on every subpopulation needs to be larger than some thresholds. The BAICS problem aims at correctly identify, with high confidence, the arm with the largest expected reward from all arms that satisfy subpopulation constraints. We analyze the complexity of the BAICS problem by proving a best achievable lower bound on the sample complexity with closed-form representation. We then design an algorithm and prove that the algorithm's sample complexity matches with the lower bound in terms of order. A brief account of numerical experiments are conducted to illustrate the theoretical findings.

Plasticity, the ability of a neural network to quickly change its predictions in response to new information, is essential for the adaptability and robustness of deep reinforcement learning systems. Deep neural networks are known to lose plasticity over the course of training even in relatively simple learning problems, but the mechanisms driving this phenomenon are still poorly understood. This paper conducts a systematic empirical analysis into plasticity loss, with the goal of understanding the phenomenon mechanistically in order to guide the future development of targeted solutions. We find that loss of plasticity is deeply connected to changes in the curvature of the loss landscape, but that it often occurs in the absence of saturated units. Based on this insight, we identify a number of parameterization and optimization design choices which enable networks to better preserve plasticity over the course of training. We validate the utility of these findings on larger-scale RL benchmarks in the Arcade Learning Environment., Comment: Accepted to ICML 2023 (oral presentation)

In biomedical science, analyzing treatment effect heterogeneity plays an essential role in assisting personalized medicine. The main goals of analyzing treatment effect heterogeneity include estimating treatment effects in clinically relevant subgroups and predicting whether a patient subpopulation might benefit from a particular treatment. Conventional approaches often evaluate the subgroup treatment effects via parametric modeling and can thus be susceptible to model mis-specifications. In this manuscript, we take a model-free semiparametric perspective and aim to efficiently evaluate the heterogeneous treatment effects of multiple subgroups simultaneously under the one-step targeted maximum-likelihood estimation (TMLE) framework. When the number of subgroups is large, we further expand this path of research by looking at a variation of the one-step TMLE that is robust to the presence of small estimated propensity scores in finite samples. From our simulations, our method demonstrates substantial finite sample improvements compared to conventional methods. In a case study, our method unveils the potential treatment effect heterogeneity of rs12916-T allele (a proxy for statin usage) in decreasing Alzheimer's disease risk., Comment: Accepted by Biometrics 2022

Constructing asymptotically valid confidence intervals through a valid central limit theorem is crucial for A/B tests, where a classical goal is to statistically assert whether a treatment plan is significantly better than a control plan. In some emerging applications for online platforms, the treatment plan is not a single plan, but instead encompasses an infinite continuum of plans indexed by a continuous treatment parameter. As such, the experimenter not only needs to provide valid statistical inference, but also desires to effectively and adaptively find the optimal choice of value for the treatment parameter to use for the treatment plan. However, we find that classical optimization algorithms, despite of their fast convergence rates under convexity assumptions, do not come with a central limit theorem that can be used to construct asymptotically valid confidence intervals. We fix this issue by providing a new optimization algorithm that on one hand maintains the same fast convergence rate and on the other hand permits the establishment of a valid central limit theorem. We discuss practical implementations of the proposed algorithm and conduct numerical experiments to illustrate the theoretical findings.

We study a game between liquidity provider and liquidity taker agents interacting in an over-the-counter market, for which the typical example is foreign exchange. We show how a suitable design of parameterized families of reward functions coupled with shared policy learning constitutes an efficient solution to this problem. By playing against each other, our deep-reinforcement-learning-driven agents learn emergent behaviors relative to a wide spectrum of objectives encompassing profit-and-loss, optimal execution and market share. In particular, we find that liquidity providers naturally learn to balance hedging and skewing, where skewing refers to setting their buy and sell prices asymmetrically as a function of their inventory. We further introduce a novel RL-based calibration algorithm which we found performed well at imposing constraints on the game equilibrium. On the theoretical side, we are able to show convergence rates for our multi-agent policy gradient algorithm under a transitivity assumption, closely related to generalized ordinal potential games.

Nonsmooth nonconvex optimization problems broadly emerge in machine learning and business decision making, whereas two core challenges impede the development of efficient solution methods with finite-time convergence guarantee: the lack of computationally tractable optimality criterion and the lack of computationally powerful oracles. The contributions of this paper are two-fold. First, we establish the relationship between the celebrated Goldstein subdifferential~\citep{Goldstein-1977-Optimization} and uniform smoothing, thereby providing the basis and intuition for the design of gradient-free methods that guarantee the finite-time convergence to a set of Goldstein stationary points. Second, we propose the gradient-free method (GFM) and stochastic GFM for solving a class of nonsmooth nonconvex optimization problems and prove that both of them can return a $(\delta,\epsilon)$-Goldstein stationary point of a Lipschitz function $f$ at an expected convergence rate at $O(d^{3/2}\delta^{-1}\epsilon^{-4})$ where $d$ is the problem dimension. Two-phase versions of GFM and SGFM are also proposed and proven to achieve improved large-deviation results. Finally, we demonstrate the effectiveness of 2-SGFM on training ReLU neural networks with the \textsc{Minst} dataset., Comment: Accepted by NeurIPS 2022; 32 pages, 18 figures; Fix a confusing part in the proof of Theorem 3.1: we use Bertsekas [1973, Proposition 2.3] rather than Bertsekas [1973, Proposition 2.4] here and do not assume the convexity of the function f

In this paper we estimate the planar Tur\'an number $\mathrm{ex}_\mathcal{P}(n,H)$ of some graphs $H$, i.e., the maximum number of edges in a planar graph $G$ of $n$ vertices not containing $H$ as a subgraph. We give a new, short proof when $H=C_5$, and study the cases when $G$ is bipartite or triangle-free and $H$ is a short even cycle. The proofs are mostly new applications or variants of the "contribution method" introduced by Ghosh, Gy\H{o}ri, Martin, Paulos and Xiao in arXiv:2004.14094.

In an influential paper, Courtois and Semal (1984) establish that when $G$ is an irreducible substochastic matrix for which $\sum_{n=0}^{\infty}G^n <\infty$, then the stationary distribution of any stochastic matrix $P\ge G$ can be expressed as a convex combination of the normalized rows of $(I-G)^{-1} = \sum_{n=0}^{\infty} G^n$. In this note, we give a short proof of this result that extends the theory to the countably infinite and continuous state space settings. This result plays an important role in obtaining error bounds in algorithms involving nearly decomposable Markov chains, and also in state truncations for Markov chains. We also use the representation to establish a new total variation distance error bound for truncated Markov chains.

Understanding the impact of the most effective policies or treatments on a response variable of interest is desirable in many empirical works in economics, statistics and other disciplines. Due to the widespread winner's curse phenomenon, conventional statistical inference assuming that the top policies are chosen independent of the random sample may lead to overly optimistic evaluations of the best policies. In recent years, given the increased availability of large datasets, such an issue can be further complicated when researchers include many covariates to estimate the policy or treatment effects in an attempt to control for potential confounders. In this manuscript, to simultaneously address the above-mentioned issues, we propose a resampling-based procedure that not only lifts the winner's curse in evaluating the best policies observed in a random sample, but also is robust to the presence of many covariates. The proposed inference procedure yields accurate point estimates and valid frequentist confidence intervals that achieve the exact nominal level as the sample size goes to infinity for multiple best policy effect sizes. We illustrate the finite-sample performance of our approach through Monte Carlo experiments and two empirical studies, evaluating the most effective policies in charitable giving and the most beneficial group of workers in the National Supported Work program., Comment: Accepted by The Journal of Econometrics

It is known that every strong tournament has directed cycles of any length, and thereby strong subtournaments of any size. In this note, we prove that they also can share a common vertex which is a king of all of them. This common vertex can be any king in the whole tournament. Further, the Hamiltonian cycles in them can be recursively constructed by inserting an additional vertex to one directed edge.

We study the stochastic multi-armed bandit problem and design new policies that enjoy both worst-case optimality for expected regret and light-tailed risk for regret distribution. Specifically, our policy design (i) enjoys the worst-case optimality for the expected regret at order $O(\sqrt{KT\ln T})$ and (ii) has the worst-case tail probability of incurring a regret larger than any $x>0$ being upper bounded by $\exp(-\Omega(x/\sqrt{KT}))$, a rate that we prove to be best achievable with respect to $T$ for all worst-case optimal policies. Our proposed policy achieves a delicate balance between doing more exploration at the beginning of the time horizon and doing more exploitation when approaching the end, compared to standard confidence-bound-based policies. We also enhance the policy design to accommodate the "any-time" setting where $T$ is unknown a priori, and prove equivalently desired policy performances as compared to the "fixed-time" setting with known $T$. Numerical experiments are conducted to illustrate the theoretical findings. We find that from a managerial perspective, our new policy design yields better tail distributions and is preferable than celebrated policies especially when (i) there is a risk of under-estimating the volatility profile, or (ii) there is a challenge of tuning policy hyper-parameters. We conclude by extending our proposed policy design to the stochastic linear bandit setting that leads to both worst-case optimality in terms of expected regret and light-tailed risk on the regret distribution., Comment: Preliminary version appeared in NeurIPS 2022

Planning with a learned model is arguably a key component of intelligence. There are several challenges in realizing such a component in large-scale reinforcement learning (RL) problems. One such challenge is dealing effectively with continuous action spaces when using tree-search planning (e.g., it is not feasible to consider every action even at just the root node of the tree). In this paper we present a method for selecting affordances useful for planning -- for learning which small number of actions/options from a continuous space of actions/options to consider in the tree-expansion process during planning. We consider affordances that are goal-and-state-conditional mappings to actions/options as well as unconditional affordances that simply select actions/options available in all states. Our selection method is gradient based: we compute gradients through the planning procedure to update the parameters of the function that represents affordances. Our empirical work shows that it is feasible to learn to select both primitive-action and option affordances, and that simultaneously learning to select affordances and planning with a learned value-equivalent model can outperform model-free RL.

We formulate selecting the best optimizing system (SBOS) problems and provide solutions for those problems. In an SBOS problem, a finite number of systems are contenders. Inside each system, a continuous decision variable affects the system's expected performance. An SBOS problem compares different systems based on their expected performances under their own optimally chosen decision to select the best, without advance knowledge of expected performances of the systems nor the optimizing decision inside each system. We design easy-to-implement algorithms that adaptively chooses a system and a choice of decision to evaluate the noisy system performance, sequentially eliminates inferior systems, and eventually recommends a system as the best after spending a user-specified budget. The proposed algorithms integrate the stochastic gradient descent method and the sequential elimination method to simultaneously exploit the structure inside each system and make comparisons across systems. For the proposed algorithms, we prove exponential rates of convergence to zero for the probability of false selection, as the budget grows to infinity. We conduct three numerical examples that represent three practical cases of SBOS problems. Our proposed algorithms demonstrate consistent and stronger performances in terms of the probability of false selection over benchmark algorithms under a range of problem settings and sampling budgets., Comment: Code in https://github.com/nian-si/SelectOptSys

Let the crown $C_{13}$ be the linear $3$-graph on $9$ vertices $\{a,b,c,d,e,f,g,h,i\}$ with edges $$E = \{\{a,b,c\}, \{a, d,e\}, \{b, f, g\}, \{c, h,i\}\}.$$ Proving a conjecture of Gy\'arf\'as et. al., we show that for any crown-free linear $3$-graph $G$ on $n$ vertices, its number of edges satisfy $$\lvert E(G) \rvert \leq \frac{3(n - s)}{2}$$ where $s$ is the number of vertices in $G$ with degree at least $6$. This result, combined with previous work, essentially completes the determination of linear Tur\'an number for linear $3$-graphs with at most $4$ edges., Comment: 5 pages, 1 figures. Correct Typos, and add acknowledgement to Professor Gyarfas

Motivated by emerging applications such as live-streaming e-commerce, promotions and recommendations, we introduce and solve a general class of non-stationary multi-armed bandit problems that have the following two features: (i) the decision maker can pull and collect rewards from up to $K\,(\ge 1)$ out of $N$ different arms in each time period; (ii) the expected reward of an arm immediately drops after it is pulled, and then non-parametrically recovers as the arm's idle time increases. With the objective of maximizing the expected cumulative reward over $T$ time periods, we design a class of ``Purely Periodic Policies'' that jointly set a period to pull each arm. For the proposed policies, we prove performance guarantees for both the offline problem and the online problems. For the offline problem when all model parameters are known, the proposed periodic policy obtains an approximation ratio that is at the order of $1-\mathcal O(1/\sqrt{K})$, which is asymptotically optimal when $K$ grows to infinity. For the online problem when the model parameters are unknown and need to be dynamically learned, we integrate the offline periodic policy with the upper confidence bound procedure to construct on online policy. The proposed online policy is proved to approximately have $\widetilde{\mathcal O}(N\sqrt{T})$ regret against the offline benchmark. Our framework and policy design may shed light on broader offline planning and online learning applications with non-stationary and recovering rewards., Comment: v1 accepted by ICML 2021