Conveying complex objectives to reinforcement learning (RL) agents often requires meticulous reward engineering. Preference-based RL methods are able to learn a more flexible reward model based on human preferences by actively incorporating human feedback, i.e. teacher's preferences between two clips of behaviors. However, poor feedback-efficiency still remains a problem in current preference-based RL algorithms, as tailored human feedback is very expensive. To handle this issue, previous methods have mainly focused on improving query selection and policy initialization. At the same time, recent exploration methods have proven to be a recipe for improving sample-efficiency in RL. We present an exploration method specifically for preference-based RL algorithms. Our main idea is to design an intrinsic reward by measuring the novelty based on learned reward. Specifically, we utilize disagreement across ensemble of learned reward models. Our intuition is that disagreement in learned reward model reflects uncertainty in tailored human feedback and could be useful for exploration. Our experiments show that exploration bonus from uncertainty in learned reward improves both feedback- and sample-efficiency of preference-based RL algorithms on complex robot manipulation tasks from MetaWorld benchmarks, compared with other existing exploration methods that measure the novelty of state visitation.
In this report, we describe a new data set called VoynaSlov which contains 21M+ Russian-language social media activities (i.e. tweets, posts, comments) made by Russian media outlets and by the general public during the time of war between Ukraine and Russia. We scraped the data from two major platforms that are widely used in Russia: Twitter and VKontakte (VK), a Russian social media platform based in Saint Petersburg commonly referred to as "Russian Facebook". We provide descriptions of our data collection process and data statistics that compare state-affiliated and independent Russian media, and also the two platforms, VK and Twitter. The main differences that distinguish our data from previously released data related to the ongoing war are its focus on Russian media and consideration of state-affiliation as well as the inclusion of data from VK, which is more suitable than Twitter for understanding Russian public sentiment considering its wide use within Russia. We hope our data set can facilitate future research on information warfare and ultimately enable the reduction and prevention of disinformation and opinion manipulation campaigns. The data set is available at https://github.com/chan0park/VoynaSlov and will be regularly updated as we continuously collect more data.
The matching principles behind optimal transport (OT) play an increasingly important role in machine learning, a trend which can be observed when OT is used to disambiguate datasets in applications (e.g. single-cell genomics) or used to improve more complex methods (e.g. balanced attention in transformers or self-supervised learning). To scale to more challenging problems, there is a growing consensus that OT requires solvers that can operate on millions, not thousands, of points. The low-rank optimal transport (LOT) approach advocated in \cite{scetbon2021lowrank} holds several promises in that regard, and was shown to complement more established entropic regularization approaches, being able to insert itself in more complex pipelines, such as quadratic OT. LOT restricts the search for low-cost couplings to those that have a low-nonnegative rank, yielding linear time algorithms in cases of interest. However, these promises can only be fulfilled if the LOT approach is seen as a legitimate contender to entropic regularization when compared on properties of interest, where the scorecard typically includes theoretical properties (statistical bounds, relation to other methods) or practical aspects (debiasing, hyperparameter tuning, initialization). We target each of these areas in this paper in order to cement the impact of low-rank approaches in computational OT.
Multi-Agent Reinforcement Learning (MARL) is useful in many problems that require the cooperation and coordination of multiple agents. Learning optimal policies using reinforcement learning in a multi-agent setting can be very difficult as the number of agents increases. Recent solutions such as Value Decomposition Networks (VDN), QMIX, QTRAN and QPLEX adhere to the centralized training and decentralized execution scheme and perform factorization of the joint action-value functions. However, these methods still suffer from increased environmental complexity, and at times fail to converge in a stable manner. We propose a novel concept of Residual Q-Networks (RQNs) for MARL, which learns to transform the individual Q-value trajectories in a way that preserves the Individual-Global-Max criteria (IGM), but is more robust in factorizing action-value functions. The RQN acts as an auxiliary network that accelerates convergence and will become obsolete as the agents reach the training objectives. The performance of the proposed method is compared against several state-of-the-art techniques such as QPLEX, QMIX, QTRAN and VDN, in a range of multi-agent cooperative tasks. The results illustrate that the proposed method, in general, converges faster, with increased stability and shows robust performance in a wider family of environments. The improvements in results are more prominent in environments with severe punishments for non-cooperative behaviours and especially in the absence of complete state information during training time.
High performance packet classification is a key component to support scalable network applications like firewalls, intrusion detection, and differentiated services. With ever increasing in the line-rate in core networks, it becomes a great challenge to design a scalable and high performance packet classification solution using hand-tuned heuristics approaches. In this paper, we present a scalable learning-based packet classification engine and its performance evaluation. By exploiting the sparsity of ruleset, our algorithm uses a few effective bits (EBs) to extract a large number of candidate rules with just a few of memory access. These effective bits are learned with deep reinforcement learning and they are used to create a bitmap to filter out the majority of rules which do not need to be full-matched to improve the online system performance. Moreover, our EBs learning-based selection method is independent of the ruleset, which can be applied to varying rulesets. Our multibit tries classification engine outperforms lookup time both in worst and average case by 55% and reduce memory footprint, compared to traditional decision tree without EBs.
The digitization of the energy infrastructure enables new, data driven, applications often supported by machine learning models. However, domain specific data transformations, pre-processing and management in modern data driven pipelines is yet to be addressed. In this paper we perform a first time study on data models, energy feature engineering and feature management solutions for developing ML-based energy applications. We first propose a taxonomy for designing data models suitable for energy applications, analyze feature engineering techniques able to transform the data model into features suitable for ML model training and finally also analyze available designs for feature stores. Using a short-term forecasting dataset, we show the benefits of designing richer data models and engineering the features on the performance of the resulting models. Finally, we benchmark three complementary feature management solutions, including an open-source feature store.
Recent works on sparse neural network training (sparse training) have shown that a compelling trade-off between performance and efficiency can be achieved by training intrinsically sparse neural networks from scratch. Existing sparse training methods usually strive to find the best sparse subnetwork possible in one single run, without involving any expensive dense or pre-training steps. For instance, dynamic sparse training (DST), as one of the most prominent directions, is capable of reaching a competitive performance of dense training by iteratively evolving the sparse topology during the course of training. In this paper, we argue that it is better to allocate the limited resources to create multiple low-loss sparse subnetworks and superpose them into a stronger one, instead of allocating all resources entirely to find an individual subnetwork. To achieve this, two desiderata are required: (1) efficiently producing many low-loss subnetworks, the so-called cheap tickets, within one training process limited to the standard training time used in dense training; (2) effectively superposing these cheap tickets into one stronger subnetwork without going over the constrained parameter budget. To corroborate our conjecture, we present a novel sparse training approach, termed \textbf{Sup-tickets}, which can satisfy the above two desiderata concurrently in a single sparse-to-sparse training process. Across various modern architectures on CIFAR-10/100 and ImageNet, we show that Sup-tickets integrates seamlessly with the existing sparse training methods and demonstrates consistent performance improvement.
We investigate the problem of co-designing computation and communication in a multi-agent system (e.g. a sensor network or a multi-robot team). We consider the realistic setting where each agent acquires sensor data and is capable of local processing before sending updates to a base station, which is in charge of making decisions or monitoring phenomena of interest in real time. Longer processing at an agent leads to more informative updates but also larger delays, giving rise to a delay-accuracy-tradeoff in choosing the right amount of local processing at each agent. We assume that the available communication resources are limited due to interference, bandwidth, and power constraints. Thus, a scheduling policy needs to be designed to suitably share the communication channel among the agents. To that end, we develop a general formulation to jointly optimize the local processing at the agents and the scheduling of transmissions. Our novel formulation leverages the notion of Age of Information to quantify the freshness of data and capture the delays caused by computation and communication. We develop efficient resource allocation algorithms using the Whittle index approach and demonstrate our proposed algorithms in two practical applications: multi-agent occupancy grid mapping in time-varying environments, and ride sharing in autonomous vehicle networks. Our experiments show that the proposed co-design approach leads to a substantial performance improvement (18-82% in our tests).
Artificial intelligence (AI) and machine learning (ML) have become vital to remain competitive for financial services companies around the globe. The two models currently competing for the pole position in credit risk management are deep learning (DL) and gradient boosting machines (GBM). This paper benchmarked those two algorithms in the context of credit scoring using three distinct datasets with different features to account for the reality that model choice/power is often dependent on the underlying characteristics of the dataset. The experiment has shown that GBM tends to be more powerful than DL and has also the advantage of speed due to lower computational requirements. This makes GBM the winner and choice for credit scoring. However, it was also shown that the outperformance of GBM is not always guaranteed and ultimately the concrete problem scenario or dataset will determine the final model choice. Overall, based on this study both algorithms can be considered state-of-the-art for binary classification tasks on structured datasets, while GBM should be the go-to solution for most problem scenarios due to easier use, significantly faster training time, and superior accuracy.
The design of recurrent neural networks (RNNs) to accurately process sequential inputs with long-time dependencies is very challenging on account of the exploding and vanishing gradient problem. To overcome this, we propose a novel RNN architecture which is based on a structure preserving discretization of a Hamiltonian system of second-order ordinary differential equations that models networks of oscillators. The resulting RNN is fast, invertible (in time), memory efficient and we derive rigorous bounds on the hidden state gradients to prove the mitigation of the exploding and vanishing gradient problem. A suite of experiments are presented to demonstrate that the proposed RNN provides state of the art performance on a variety of learning tasks with (very) long time-dependencies.