We study fair mechanisms for the (asymmetric) one-sided allocation problem with m items and n multi-unit demand agents with additive, unit-sum valuations. The symmetric case (m=n), the one-sided matching problem, has been studied extensively for the class of unit demand agents, in particular with respect to the folklore Random Priority mechanism and the Probabilistic Serial mechanism, introduced by Bogomolnaia and Moulin. Under the assumption of unit-sum valuation functions, Christodoulou et al. proved that the price of anarchy is $\Theta(\sqrt{n})$ in the one-sided matching problem for both the Random Priority and Probabilistic Serial mechanisms. Whilst both Random Priority and Probabilistic Serial are ordinal mechanisms, these approximation guarantees are the best possible even for the broader class of cardinal mechanisms. To extend these results to the general setting there are two technical obstacles. One, asymmetry ($m\neq n$) is problematic especially when the number of items is much greater than the number of items. Two, it is necessary to study multi-unit demand agents rather than simply unit demand agents. Our approach is to study a cardinal mechanism variant of Probabilistic Serial, which we call Cardinal Probabilistic Serial. We present structural theorems for this mechanism and use them to obtain bounds on the price of anarchy. Our first main result is an upper bound of $O(\sqrt{n}\cdot \log m)$ on the price of anarchy for the asymmetric one-sided allocation problem with multi-unit demand agents. This upper bound applies to Probabilistic Serial as well and there is a complementary lower bound of $\Omega(\sqrt{n})$ for any fair mechanism. Our second main result is that the price of anarchy degrades with the number of items. Specifically, a logarithmic dependence on the number of items is necessary for both mechanisms.
相關內容
Software testing activities scrutinize the artifacts and the behavior of a software product to find possible defects and ensure that the product meets its expected requirements. Recently, Deep Reinforcement Learning (DRL) has been successfully employed in complex testing tasks such as game testing, regression testing, and test case prioritization to automate the process and provide continuous adaptation. Practitioners can employ DRL by implementing from scratch a DRL algorithm or using a DRL framework. DRL frameworks offer well-maintained implemented state-of-the-art DRL algorithms to facilitate and speed up the development of DRL applications. Developers have widely used these frameworks to solve problems in various domains including software testing. However, to the best of our knowledge, there is no study that empirically evaluates the effectiveness and performance of implemented algorithms in DRL frameworks. Moreover, some guidelines are lacking from the literature that would help practitioners choose one DRL framework over another. In this paper, we empirically investigate the applications of carefully selected DRL algorithms on two important software testing tasks: test case prioritization in the context of Continuous Integration (CI) and game testing. For the game testing task, we conduct experiments on a simple game and use DRL algorithms to explore the game to detect bugs. Results show that some of the selected DRL frameworks such as Tensorforce outperform recent approaches in the literature. To prioritize test cases, we run experiments on a CI environment where DRL algorithms from different frameworks are used to rank the test cases. Our results show that the performance difference between implemented algorithms in some cases is considerable, motivating further investigation.
An approach is introduced for comparing the estimated states of stochastic compartmental models for an epidemic or biological process with analytically obtained solutions from the corresponding system of ordinary differential equations (ODEs). Positive integer valued samples from a stochastic model are generated numerically at discrete time intervals using either the Reed-Frost chain Binomial or Gillespie algorithm. The simulated distribution of realisations is compared with an exact solution obtained analytically from the ODE model. Using this novel methodology this work demonstrates it is feasible to check that the realisations from the stochastic compartmental model adhere to the ODE model they represent. There is no requirement for the model to be in any particular state or limit. These techniques are developed using the stochastic compartmental model for a susceptible-infected-recovered (SIR) epidemic process. The Lotka-Volterra model is then used as an example of the generality of the principles developed here. This approach presents a way of testing/benchmarking the numerical solutions of stochastic compartmental models, e.g. using unit tests, to check that the computer code along with its corresponding algorithm adheres to the underlying ODE model.
Human supervisors in multi-robot systems are primarily responsible for monitoring robots, but can also be assigned with secondary tasks. These tasks can act as interruptions and can be categorized as either intrinsic, i.e., being directly related to the monitoring task, or extrinsic, i.e., being unrelated. In this paper, we investigate the impact of these two types of interruptions through a user study ($N=39$), where participants monitor a number of remote mobile robots while intermittently being interrupted by either a robot fault correction task (intrinsic) or a messaging task (extrinsic). We find that task performance of participants does not change significantly with the interruptions but depends greatly on the number of robots. However, interruptions result in an increase in perceived workload, and extrinsic interruptions have a more negative effect on workload across all NASA-TLX scales. Participants also reported switching between extrinsic interruptions and the primary task to be more difficult compared to the intrinsic interruption case. Statistical significance of these results is confirmed using ANOVA and one-sample t-test. These findings suggest that when deciding task assignment in such supervision systems, one should limit interruptions from secondary tasks, especially extrinsic ones, in order to limit user workload.
Mathematical Selection is a method in which we select a particular choice from a set of such. It have always been an interesting field of study for mathematicians. Accordingly, Combinatorial Optimization is a sub field of this domain of Mathematical Selection, where we generally, deal with problems subjecting to Operation Research, Artificial Intelligence and many more promising domains. In a broader sense, an optimization problem entails maximising or minimising a real function by systematically selecting input values from within an allowed set and computing the function's value. A broad region of applied mathematics is the generalisation of metaheuristic theory and methods to other formulations. More broadly, optimization entails determining the finest virtues of some fitness function, offered a fixed space, which may include a variety of distinct types of decision variables and contexts. In this work, we will be working on the famous Balanced Assignment Problem, and will propose a comparative analysis on the Complexity Metrics of Computational Time for different Notions of solving the Balanced Assignment Problem.
Despite the popularity and success of deep learning, there is limited understanding of when, how, and why neural networks generalize to unseen examples. Since learning can be seen as extracting information from data, we formally study information captured by neural networks during training. Specifically, we start with viewing learning in presence of noisy labels from an information-theoretic perspective and derive a learning algorithm that limits label noise information in weights. We then define a notion of unique information that an individual sample provides to the training of a deep network, shedding some light on the behavior of neural networks on examples that are atypical, ambiguous, or belong to underrepresented subpopulations. We relate example informativeness to generalization by deriving nonvacuous generalization gap bounds. Finally, by studying knowledge distillation, we highlight the important role of data and label complexity in generalization. Overall, our findings contribute to a deeper understanding of the mechanisms underlying neural network generalization.
This paper proposes a new one-sided matching market model in which every agent has a cost function that is allowed to take a negative value. Our model aims to capture the situation where some agents can profit by exchanging their obtained goods with other agents. We formulate such a model based on a graphical one-sided matching market, introduced by Massand and Simon [Massand and Simon, IJCAI 2019]. We investigate the existence of stable outcomes for such a market. We prove that there is an instance that has no core-stable allocation. On the other hand, we guarantee the existence of two-stable allocations even where exchange costs exist. However, it is PLS-hard to find a two-stable allocation for a market with exchange costs even if the maximum degree of the graph is five.
Classification of unlabeled data is usually achieved by supervised learning from labeled samples. Although there exist many sophisticated supervised machine learning methods that can predict the missing labels with a high level of accuracy, they often lack the required transparency in situations where it is important to provide interpretable results and meaningful measures of confidence. Body fluid classification of forensic casework data is the case in point. We develop a new Biclustering Dirichlet Process (BDP), with a three-level hierarchy of clustering, and a model-based approach to classification which adapts to block structure in the data matrix. As the class labels of some observations are missing, the number of rows in the data matrix for each class is unknown. The BDP handles this and extends existing biclustering methods by simultaneously biclustering multiple matrices each having a randomly variable number of rows. We demonstrate our method by applying it to the motivating problem, which is the classification of body fluids based on mRNA profiles taken from crime scenes. The analyses of casework-like data show that our method is interpretable and produces well-calibrated posterior probabilities. Our model can be more generally applied to other types of data with a similar structure to the forensic data.
Context. Algorithmic racism is the term used to describe the behavior of technological solutions that constrains users based on their ethnicity. Lately, various data-driven software systems have been reported to discriminate against Black people, either for the use of biased data sets or due to the prejudice propagated by software professionals in their code. As a result, Black people are experiencing disadvantages in accessing technology-based services, such as housing, banking, and law enforcement. Goal. This study aims to explore algorithmic racism from the perspective of software professionals. Method. A survey questionnaire was applied to explore the understanding of software practitioners on algorithmic racism, and data analysis was conducted using descriptive statistics and coding techniques. Results. We obtained answers from a sample of 73 software professionals discussing their understanding and perspectives on algorithmic racism in software development. Our results demonstrate that the effects of algorithmic racism are well-known among practitioners. However, there is no consensus on how the problem can be effectively addressed in software engineering. In this paper, some solutions to the problem are proposed based on the professionals' narratives. Conclusion. Combining technical and social strategies, including training on structural racism for software professionals, is the most promising way to address the algorithmic racism problem and its effects on the software solutions delivered to our society.
This book develops an effective theory approach to understanding deep neural networks of practical relevance. Beginning from a first-principles component-level picture of networks, we explain how to determine an accurate description of the output of trained networks by solving layer-to-layer iteration equations and nonlinear learning dynamics. A main result is that the predictions of networks are described by nearly-Gaussian distributions, with the depth-to-width aspect ratio of the network controlling the deviations from the infinite-width Gaussian description. We explain how these effectively-deep networks learn nontrivial representations from training and more broadly analyze the mechanism of representation learning for nonlinear models. From a nearly-kernel-methods perspective, we find that the dependence of such models' predictions on the underlying learning algorithm can be expressed in a simple and universal way. To obtain these results, we develop the notion of representation group flow (RG flow) to characterize the propagation of signals through the network. By tuning networks to criticality, we give a practical solution to the exploding and vanishing gradient problem. We further explain how RG flow leads to near-universal behavior and lets us categorize networks built from different activation functions into universality classes. Altogether, we show that the depth-to-width ratio governs the effective model complexity of the ensemble of trained networks. By using information-theoretic techniques, we estimate the optimal aspect ratio at which we expect the network to be practically most useful and show how residual connections can be used to push this scale to arbitrary depths. With these tools, we can learn in detail about the inductive bias of architectures, hyperparameters, and optimizers.
Current deep learning research is dominated by benchmark evaluation. A method is regarded as favorable if it empirically performs well on the dedicated test set. This mentality is seamlessly reflected in the resurfacing area of continual learning, where consecutively arriving sets of benchmark data are investigated. The core challenge is framed as protecting previously acquired representations from being catastrophically forgotten due to the iterative parameter updates. However, comparison of individual methods is nevertheless treated in isolation from real world application and typically judged by monitoring accumulated test set performance. The closed world assumption remains predominant. It is assumed that during deployment a model is guaranteed to encounter data that stems from the same distribution as used for training. This poses a massive challenge as neural networks are well known to provide overconfident false predictions on unknown instances and break down in the face of corrupted data. In this work we argue that notable lessons from open set recognition, the identification of statistically deviating data outside of the observed dataset, and the adjacent field of active learning, where data is incrementally queried such that the expected performance gain is maximized, are frequently overlooked in the deep learning era. Based on these forgotten lessons, we propose a consolidated view to bridge continual learning, active learning and open set recognition in deep neural networks. Our results show that this not only benefits each individual paradigm, but highlights the natural synergies in a common framework. We empirically demonstrate improvements when alleviating catastrophic forgetting, querying data in active learning, selecting task orders, while exhibiting robust open world application where previously proposed methods fail.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191