IA Scholar Query: On the scaling of polynomial features for representation matching.
https://scholar.archive.org/
Internet Archive Scholar query results feedeninfo@archive.orgThu, 15 Dec 2022 00:00:00 GMTfatcat-scholarhttps://scholar.archive.org/help1440Efficient assimilation of sparse data into RANS-based turbulent flow simulations using a discrete adjoint method
https://scholar.archive.org/work/vh4pdghs6nhzzmnuoavfqfba7q
Turbulent flow simulations based on the Reynolds-averaged Navier-Stokes (RANS) equations continue to be the workhorse approach for industrial flow problems. However, due to the inherent averaging and the closure models required for the Reynolds stresses, their accuracy is limited. Experimental data, on the other hand, represents the true flow features but is potentially only available in limited locations or at low resolution. Data assimilation (DA) can be used to combine closure models with such experimental data to obtain accurate simulation results. In this scenario, the result of the DA process can be interpreted as a physics-based interpolation and serve as a basis for further studies of the flow. Moreover, such tuned models may be used in machine learning approaches aiming at a priori closure model enhancement. The main objective of this work is to recover a spatially varying eddy viscosity correction factor from sparsely distributed reference data. We use an efficient in-house data assimilation implementation based on the discrete adjoint method for RANS simulations in OpenFOAM to recover an optimal eddy viscosity field. A gradient optimization procedure is used to minimize a velocity-based cost function with a regularization term. All the partial derivatives in the gradient computation are evaluated with an efficient semi-analytical approach at the cost of approximately one forward solution step. The effectiveness of the proposed approach is demonstrated for a periodic hill test case at Re = 10595 with different spatial distributions of the reference data. The results show improvements in the agreement between simulation and reference for a range of reference data configurations and highlight the performance of linear eddy viscosity models.Oliver Brenner, Pasha Piroozmand, Patrick Jennywork_vh4pdghs6nhzzmnuoavfqfba7qThu, 15 Dec 2022 00:00:00 GMTFusion of Gabor filter and steerable pyramid to improve iris recognition system
https://scholar.archive.org/work/4ldiofabp5h6xf5xyggks452ty
<span>Iris recognition system is a technique of identifying people using their distinctive features. Generally, this technique is used in security, because it offers a good reliability. Different researchers have proposed new methods for iris recognition system to increase its effectiveness. In this paper, we propose a new method for iris recognition based on Gabor filter and steerable pyramid decomposition. It's an efficient and accurate linear multi-scale, multi-orientation image decomposition to capture texture details of an image. At first, the iris image is segmented, normalized and decomposed by Gabor filter and steerable pyramid method. Multiple sub-band are generated by applying steerable pyramid on the input image. High frequency sub-band is ignored to eliminate noise and increase the accuracy. The method was validated using CASIA-v4 (Chinese Academy of Sciences Institute of Automation), IITD (</span><span>Indian Institute of Technology Delhi) and UPOL (University of Phoenix Online) databases. The performance of the proposed method is better than the most methods in the literature. The proposed algorithm provides accuracy of 99.99%. False acceptance rate (FAR), equal error rate (EER) and genuine acceptance rate (GAR) have also been improved.</span>Mohamed Radouane, Nadia Idrissi Zouggari, Amine Amraoui, Mounir Amraouiwork_4ldiofabp5h6xf5xyggks452tyThu, 01 Dec 2022 00:00:00 GMTBrain white matter hyperintensity lesion characterization in 3D T2 fluid-attenuated inversion recovery magnetic resonance images: Shape, texture, and their correlations with potential growth
https://scholar.archive.org/work/ax3ikdstfrbqxmryfk3haibkue
Analyses of age-related white matter hyperintensity (WMH) lesions manifested in T2 fluid-attenuated inversion recovery (FLAIR) magnetic resonance images (MRI) have been mostly on understanding the size and location of the WMH lesions and rarely on the morphological characterization of the lesions. This work extends our prior analyses of the morphological characteristics and texture of WMH from 2D to 3D based on 3D T2 FLAIR images. 3D Zernike transformation was used to characterize WMH shape; a fuzzy logic method was used to characterize the lesion texture. We then clustered 3D WMH lesions into groups based on their 3D shape and texture features. A potential growth index (PGI) to assess dynamic changes in WMH lesions was developed based on the image texture features of the WMH lesion penumbra. WMH lesions with various sizes were segmented from brain images of 32 cognitively normal older adults. The WMH lesions were divided into two groups based on their size. Analyses of Variance (ANOVAs) showed significant differences in PGI among WMH shape clusters (P = 1.57 × 10–3 for small lesions; P = 3.14 × 10–2 for large lesions). Significant differences in PGI were also found among WMH texture group clusters (P = 1.79 × 10–6). In conclusion, we presented a novel approach to characterize the morphology of 3D WMH lesions and explored the potential to assess the dynamic morphological changes of WMH lesions using PGI.Chih-Ying Gwo, David C. Zhu, Rong Zhangwork_ax3ikdstfrbqxmryfk3haibkueThu, 24 Nov 2022 00:00:00 GMTEEG Channel Selection Techniques in Motor Imagery Applications: A Review and New Perspectives
https://scholar.archive.org/work/zsbinb5muzesjciw25646j76da
Communication, neuro-prosthetics, and environmental control are just a few applications for disabled persons who use robots and manipulators that use brain-computer interface (BCI) systems. The brain's motor imagery (MI) signal is an essential input for a brain-related task in BCI applications. Due to their noninvasive, portability, and cost-effectiveness, electroencephalography (EEG) signals are the most widely used input in BCI systems. The EEG data are often collected from more than 100 different locations in the brain; channel selection techniques are critical for selecting the optimum channels for a given application. However, when analyzing EEG data, the principal purpose of channel selection is to reduce computational complexity, improve classification accuracy by avoiding overfitting, and reduce setup time. Several channel selection assessment algorithms, both with and without classification-based methods, extracted appropriate channel subsets using defined criteria. Therefore, based on the exhaustive analysis of the EEG channel selection, this manuscript analyses several existing studies to reduce the number of noisy channels and improve system performance. We review several existing works to find the most promising MI-based EEG channel selection algorithms and associated classification methodologies on various datasets. Moreover, we focus on channel selection methods that choose fewer channels with great precision. Finally, our main finding is that a smaller channel set, typically 10–30% of total channels, provided excellent performance compared to other existing studies.Abdullah, Ibrahima Faye, Md Rafiqul Islamwork_zsbinb5muzesjciw25646j76daThu, 24 Nov 2022 00:00:00 GMTUsing a quantum computer to solve a real-world problem – what can be achieved today?
https://scholar.archive.org/work/2ml5jshzzrdelpgay4qu5g5mbi
Quantum computing is an important developing technology with the potential to revolutionise the landscape of scientific and business problems that can be practically addressed. The widespread excitement derives from the potential for a fault tolerant quantum computer to solve previously intractable problems. Such a machine is not expected to be available until 2030 at least. Thus we are currently in the so-called NISQ era where more heuristic quantum approaches are being applied to early versions of quantum hardware. In this paper we seek to provide a more accessible explanation of many of the more technical aspects of quantum computing in the current NISQ era exploring the 2 main hybrid classical-quantum algorithms, QAOA and VQE, as well as quantum annealing. We apply these methods, to an example of combinatorial optimisation in the form of a facilities location problem. Methods explored include the applications of different types of mixer (X, XY and a novel 3XY mixer) within QAOA as well as the effects of many settings for important meta parameters, which are often not focused on in research papers. Similarly, we explore alternative parameter settings in the context of quantum annealing. Our research confirms the broad consensus that quantum gate hardware will need to be much more capable than is available currently in terms of scale and fidelity to be able to address such problems at a commercially valuable level. Quantum annealing is closer to offering quantum advantage but will also need to achieve a significant step up in scale and connectivity to address optimisation problems where classical solutions are sub-optimal.R.Cumming, T.Thomaswork_2ml5jshzzrdelpgay4qu5g5mbiWed, 23 Nov 2022 00:00:00 GMTImproved Bounds on Neural Complexity for Representing Piecewise Linear Functions
https://scholar.archive.org/work/a4wyk3yls5gqhhaxlqjmowbdla
A deep neural network using rectified linear units represents a continuous piecewise linear (CPWL) function and vice versa. Recent results in the literature estimated that the number of neurons needed to exactly represent any CPWL function grows exponentially with the number of pieces or exponentially in terms of the factorial of the number of distinct linear components. Moreover, such growth is amplified linearly with the input dimension. These existing results seem to indicate that the cost of representing a CPWL function is expensive. In this paper, we propose much tighter bounds and establish a polynomial time algorithm to find a network satisfying these bounds for any given CPWL function. We prove that the number of hidden neurons required to exactly represent any CPWL function is at most a quadratic function of the number of pieces. In contrast to all previous results, this upper bound is invariant to the input dimension. Besides the number of pieces, we also study the number of distinct linear components in CPWL functions. When such a number is also given, we prove that the quadratic complexity turns into bilinear, which implies a lower neural complexity because the number of distinct linear components is always not greater than the minimum number of pieces in a CPWL function. When the number of pieces is unknown, we prove that, in terms of the number of distinct linear components, the neural complexity of any CPWL function is at most polynomial growth for low-dimensional inputs and factorial growth for the worst-case scenario, which are significantly better than existing results in the literature.Kuan-Lin Chen, Harinath Garudadri, Bhaskar D. Raowork_a4wyk3yls5gqhhaxlqjmowbdlaWed, 23 Nov 2022 00:00:00 GMTLifshitz transition in the phase diagram of two-leg t-J ladder systems at low filling
https://scholar.archive.org/work/onpibnh3njcwboqfhpyrlbajom
We use a combination of numerical matrix product states (MPS) and analytical approaches to investigate the phase diagram of the two-leg t-J ladder in the region of low to intermediate fillings. We choose the same coupling strength along the leg- and rung-directions, but study the effect of adding a nearest-neighbor repulsion V. We observe a rich phase diagram and analytically identify a Lifshitz-like band filling transition, which can be associated to a numerically observed crossover from s-wave to d-wave like superconducting quasi-long range order (QLRO). Due to the strong interactions, the Lifshitz transition is smeared into a crossover region which separates two distinct Luttinger theories with unequal physical meaning of the Luttinger parameter. Our numerically exact MPS results spotlight deviations from standard Luttinger theory in this crossover region and is consistent with Luttinger theory sufficiently far away from the Lifshitz transition. At very low fillings, studying the Friedel-like oscillations of the local density identifies a precursor region to a Wigner crystal at small values of the magnetic exchange interaction J/t. We discuss analytically how tuning parameters at these fillings modifies the phase diagram, and find good agreement with MPS results.Steffen Bollmann, Alexander Osterkorn, Elio J. König, Salvatore R. Manmanawork_onpibnh3njcwboqfhpyrlbajomWed, 23 Nov 2022 00:00:00 GMTLarge hyperbolic circles
https://scholar.archive.org/work/tbylvxz6q5gzxgv4biuo5fow24
We consider circles of common centre and increasing radius on a compact hyperbolic surface and, more generally, on its unit tangent bundle. We establish a precise asymptotics for their rate of equidistribution. Our result holds for translates of any circle arc by arbitrary elements of SL_2(ℝ). Our proof relies on a spectral method pioneered by Ratner and subsequently developed by Burger in the study of geodesic and horocycle flows. We further derive statistical limit theorems, with compactly supported limiting distribution, for appropriately rescaled circle averages of sufficient regular observables. Finally, we discuss applications to the classical circle problem in the hyperbolic plane, following the approach of Duke-Rudnick-Sarnak and Eskin-McMullen.Emilio Corso, Davide Ravottiwork_tbylvxz6q5gzxgv4biuo5fow24Wed, 23 Nov 2022 00:00:00 GMTDiscretization Invariant Learning on Neural Fields
https://scholar.archive.org/work/io5orwjckfdw3gou4cwxgzrbe4
While neural fields have emerged as powerful representations of continuous data, there is a need for neural networks that can perform inference on such data without being sensitive to how the field is sampled, a property called discretization invariance. We develop DI-Net, a framework for learning discretization invariant operators on neural fields of any type. Whereas current theoretical analyses of discretization invariant networks are restricted to the limit of infinite samples, our analysis does not require infinite samples and establishes upper bounds on the variation in DI-Net outputs given different finite discretizations. Our framework leads to a family of neural networks driven by numerical integration via quasi-Monte Carlo sampling with discretizations of low discrepancy. DI-Nets manifest desirable theoretical properties such as universal approximation of a large class of maps between L^2 functions, and gradients that are also discretization invariant. DI-Nets can also be seen as generalizations of many existing network families as they bridge discrete and continuous network classes, such as convolutional neural networks (CNNs) and neural operators respectively. Experimentally, DI-Nets derived from CNNs can learn to classify and segment visual data represented by neural fields under various discretizations, and sometimes even generalize to new types of discretizations at test time. Code: https://github.com/clintonjwang/DI-net.Clinton J. Wang, Polina Gollandwork_io5orwjckfdw3gou4cwxgzrbe4Wed, 23 Nov 2022 00:00:00 GMTLow-temperature Ising dynamics with random initializations
https://scholar.archive.org/work/rjti6ofkhja33idq6wx6xznzze
It is well known that Glauber dynamics on spin systems typically suffer exponential slowdowns at low temperatures. This is due to the emergence of multiple metastable phases in the state space, separated by narrow bottlenecks that are hard for the dynamics to cross. It is a folklore belief that if the dynamics is initialized from an appropriate random mixture of ground states, one for each phase, then convergence to the Gibbs distribution should be much faster. However, such phenomena have largely evaded rigorous analysis, as most tools in the study of Markov chain mixing times are tailored to worst-case initializations. In this paper we develop a general framework towards establishing this conjectured behavior for the Ising model. In the classical setting of the Ising model on an N-vertex torus in ℤ^d, our framework implies that the mixing time for the Glauber dynamics, initialized from a 1/2-1/2 mixture of the all-plus and all-minus configurations, is N^1+o(1) in dimension d=2, and at most quasi-polynomial in all dimensions d≥ 3, at all temperatures below the critical one. The key innovation in our analysis is the introduction of the notion of "weak spatial mixing within a phase", a low-temperature adaptation of the classical concept of weak spatial mixing. We show both that this new notion is strong enough to control the mixing time from the above random initialization (by relating it to the mixing time with plus boundary condition at O(log N) scales), and that it holds at all low temperatures in all dimensions. This framework naturally extends to much more general families of graphs. To illustrate this, we also use the same approach to establish optimal O(Nlog N) mixing for the Ising Glauber dynamics on random regular graphs at sufficiently low temperatures, when initialized from the same random mixture.Reza Gheissari, Alistair Sinclairwork_rjti6ofkhja33idq6wx6xznzzeWed, 23 Nov 2022 00:00:00 GMTPrivacy-Preserving Application-to-Application Authentication Using Dynamic Runtime Behaviors
https://scholar.archive.org/work/zkldmph7c5en7erz6tonl3jocy
Application authentication is typically performed using some form of secret credentials such as cryptographic keys, passwords, or API keys. Since clients are responsible for securely storing and managing the keys, this approach is vulnerable to attacks on clients. Similarly a centrally managed key store is also susceptible to various attacks and if compromised, can leak credentials. To resolve such issues, we propose an application authentication, where we rely on unique and distinguishable application's behavior to lock the key during a setup phase and unlock it for authentication. Our system add a fuzzy-extractor layer on top of current credential authentication systems. During a key enrollment process, the application's behavioral data collected from various sensors in the network are used to hide the credential key. The fuzzy extractor releases the key to the server if the application's behavior during the authentication matches the one collected during the enrollment, with some noise tolerance. We designed the system, analyzed its security, and implemented and evaluated it using 10 real-life applications deployed in our network. Our security analysis shows that the system is secure against client compromise, vault compromise, and feature observation. The evaluation shows the scheme can achieve 0 percent False Accept Rate with an average False Rejection Rate 14 percent and takes about 51 ms to successfully authenticate a client. In light of these promising results, we expect our system to be of practical use, since its deployment requires zero to minimal changes on the server.Mihai Christodorescu, Maliheh Shirvanian, Shams Zawoadwork_zkldmph7c5en7erz6tonl3jocyWed, 23 Nov 2022 00:00:00 GMTFoundational Semantics of Dynamically Scheduled Attribute Grammar Evaluation
https://scholar.archive.org/work/anqgd2b7unas5payhfzeamv6ou
The similarities and differences between attribute grammar systems are obscured by their implementations. A formalism that captures the essence of such systems would allow for equivalence, correctness, and other analyses to be formally framed and proven. We present Saiga, a core language and small-step operational semantics that precisely captures the fundamental concepts of the evaluation of dynamically scheduled attribute grammars. We also present and discuss evaluation semantics for reference, parameterised, cached, and higher order attribute grammars. Saiga's utility is demonstrated through proofs about the system's operation, equivalence proofs between distinct Saiga attribute grammar programs, and "step count" comparisons between such programs. The language, semantics and proofs have been mechanised in Lean.Scott Buckleywork_anqgd2b7unas5payhfzeamv6ouWed, 23 Nov 2022 00:00:00 GMTVectorized MATLAB Implementation of the Incremental Minimization Principle for Rate-Independent Dissipative Solids Using FEM: A Constitutive Model of Shape Memory Alloys
https://scholar.archive.org/work/wla76xy4kjdqbl2fj7wion3njq
The incremental energy minimization principle provides a compact variational formulation for evolutionary boundary problems based on constitutive models of rate-independent dissipative solids. In this work, we develop and implement a versatile computational tool for the resolution of these problems via the finite element method (FEM). The implementation is coded in the MATLAB programming language and benefits from vector operations, allowing all local energy contributions to be evaluated over all degrees of freedom at once. The monolithic solution scheme combined with gradient-based optimization methods is applied to the inherently nonlinear, non-smooth convex minimization problem. An advanced constitutive model for shape memory alloys, which features a strongly coupled rate-independent dissipation function and several constraints on internal variables, is implemented as a benchmark example. Numerical simulations demonstrate the capabilities of the computational tool, which is suited for the rapid development and testing of advanced constitutive laws of rate-independent dissipative solids.Miroslav Frost, Jan Valdmanwork_wla76xy4kjdqbl2fj7wion3njqWed, 23 Nov 2022 00:00:00 GMTChaos in Matrix Gauge Theories with Massive Deformations
https://scholar.archive.org/work/6v2gp34s7rfdjc7v4ahzvnaxdy
Starting from an 𝑆𝑈 (𝑁) matrix quantum mechanics model with massive deformation terms and by introducing an ansatz configuration involving fuzzy four-and two-spheres with collective time dependence, we obtain a family of effective Hamiltonians, 𝐻 𝑛 , (𝑁 = 1 6 (𝑛 + 1) (𝑛 + 2) (𝑛 + 3)) and examine their emerging chaotic dynamics. Through numerical work, we model the variation of the largest Lyapunov exponents as a function of the energy and find that they vary either as ∝ (𝐸 − (𝐸 𝑛 ) 𝐹 ) 1/4 or ∝ 𝐸 1/4 , where (𝐸 𝑛 ) 𝐹 stand for the energies of the unstable fixed points of the phase space. We use our results to put upper bounds on the temperature above which the Lyapunov exponents comply with the Maldacena-Shenker-Stanford (MSS) bound, 2𝜋𝑇, and below which it will eventually be violated.Seckin Kurkcuoglu, K. Baskan, O. Oktay, C. Tasciwork_6v2gp34s7rfdjc7v4ahzvnaxdyWed, 23 Nov 2022 00:00:00 GMTDeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting
https://scholar.archive.org/work/stlrhpwnsfebreqecopkb5kv4q
End-to-end text spotting aims to integrate scene text detection and recognition into a unified framework. Dealing with the relationship between the two sub-tasks plays a pivotal role in designing effective spotters. Although transformer-based methods eliminate the heuristic post-processing, they still suffer from the synergy issue between the sub-tasks and low training efficiency. In this paper, we present DeepSolo, a simple detection transformer baseline that lets a single Decoder with Explicit Points Solo for text detection and recognition simultaneously. Technically, for each text instance, we represent the character sequence as ordered points and model them with learnable explicit point queries. After passing a single decoder, the point queries have encoded requisite text semantics and locations and thus can be further decoded to the center line, boundary, script, and confidence of text via very simple prediction heads in parallel, solving the sub-tasks in text spotting in a unified framework. Besides, we also introduce a text-matching criterion to deliver more accurate supervisory signals, thus enabling more efficient training. Quantitative experiments on public benchmarks demonstrate that DeepSolo outperforms previous state-of-the-art methods and achieves better training efficiency. In addition, DeepSolo is also compatible with line annotations, which require much less annotation cost than polygons. The code will be released.Maoyuan Ye, Jing Zhang, Shanshan Zhao, Juhua Liu, Tongliang Liu, Bo Du, Dacheng Taowork_stlrhpwnsfebreqecopkb5kv4qWed, 23 Nov 2022 00:00:00 GMTHKLL for the Non-Normalizable Mode
https://scholar.archive.org/work/5w4dr7zc3vgydaorkytpbwzkqe
We discuss various aspects of HKLL bulk reconstruction for the free scalar field in AdS_d+1. First, we consider the spacelike reconstruction kernel for the non-normalizable mode in global coordinates. We construct it as a mode sum. In even bulk dimensions, this can be reproduced using a chordal Green's function approach that we propose. This puts the global AdS results for the non-normalizable mode on an equal footing with results in the literature for the normalizable mode. In Poincaré AdS, we present explicit mode sum results in general even and odd dimensions for both normalizable and non-normalizable kernels. For generic scaling dimension Δ, these can be re-written in a form that matches with the global AdS results via an antipodal mapping, plus a remainder. We are not aware of a general argument in the literature for dropping these remainder terms, but we note that a slight complexification of a boundary spatial coordinate (which we call an i ϵ prescription) allows us to do so in cases where Δ is (half-) integer. Since the non-normalizable mode turns on a source in the CFT, our primary motivation for considering it is as a step towards understanding linear wave equations in general spacetimes from a holographic perspective. But when the scaling dimension Δ is in the Breitenlohner-Freedman window, we note that the construction has some interesting features within AdS/CFT.Budhaditya Bhattacharjee, Chethan Krishnan, Debajyoti Sarkarwork_5w4dr7zc3vgydaorkytpbwzkqeWed, 23 Nov 2022 00:00:00 GMTOn spectrally flowed local vertex operators in AdS$_3$
https://scholar.archive.org/work/4xtuo47e7bbkfadfng4fmmjpoq
We provide a novel local definition for spectrally flowed vertex operators in the SL(2,\mathbb{R})SL(2,ℝ)-WZW model, generalising the proposal of Maldacena and Ooguri in [arXiv:hep-th/0111180] for the singly-flowed case to all \omega>1ω>1. This allows us to establish the precise connection between the computation of correlators using the so-called spectral flow operator, and the methods introduced recently by Dei and Eberhardt in [arXiv:2105.12130] based on local Ward identities. We show that the auxiliary variable yy used in the latter paper arises naturally from a point-splitting procedure in the space-time coordinate. The recursion relations satisfied by spectrally flowed correlators, which take the form of partial differential equations in yy-space, then correspond to null-state conditions for generalised spectral flowed operators. We highlight the role of certain SL(2,\mathbb{R})SL(2,ℝ) discrete module isomorphisms in this context, and prove the validity of the conjecture put forward in [arXiv:2105.12130] for yy-space structure constants of three-point functions with arbitrary spectral flow charges.Sergio Iguri, Nicolas Kovenskywork_4xtuo47e7bbkfadfng4fmmjpoqWed, 23 Nov 2022 00:00:00 GMTFairness Increases Adversarial Vulnerability
https://scholar.archive.org/work/auz5icsh7rch7eqizvahnpfnyu
The remarkable performance of deep learning models and their applications in consequential domains (e.g., facial recognition) introduces important challenges at the intersection of equity and security. Fairness and robustness are two desired notions often required in learning models. Fairness ensures that models do not disproportionately harm (or benefit) some groups over others, while robustness measures the models' resilience against small input perturbations. This paper shows the existence of a dichotomy between fairness and robustness, and analyzes when achieving fairness decreases the model robustness to adversarial samples. The reported analysis sheds light on the factors causing such contrasting behavior, suggesting that distance to the decision boundary across groups as a key explainer for this behavior. Extensive experiments on non-linear models and different architectures validate the theoretical findings in multiple vision domains. Finally, the paper proposes a simple, yet effective, solution to construct models achieving good tradeoffs between fairness and robustness.Cuong Tran, Keyu Zhu, Ferdinando Fioretto, Pascal Van Hentenryckwork_auz5icsh7rch7eqizvahnpfnyuWed, 23 Nov 2022 00:00:00 GMTOn Instance-Dependent Bounds for Offline Reinforcement Learning with Linear Function Approximation
https://scholar.archive.org/work/z5uyisimvvdjrhnq5wbdgc55xu
Sample-efficient offline reinforcement learning (RL) with linear function approximation has recently been studied extensively. Much of prior work has yielded the minimax-optimal bound of 𝒪̃(1/√(K)), with K being the number of episodes in the offline data. In this work, we seek to understand instance-dependent bounds for offline RL with function approximation. We present an algorithm called Bootstrapped and Constrained Pessimistic Value Iteration (BCP-VI), which leverages data bootstrapping and constrained optimization on top of pessimism. We show that under a partial data coverage assumption, that of concentrability with respect to an optimal policy, the proposed algorithm yields a fast rate of 𝒪̃(1/K) for offline RL when there is a positive gap in the optimal Q-value functions, even when the offline data were adaptively collected. Moreover, when the linear features of the optimal actions in the states reachable by an optimal policy span those reachable by the behavior policy and the optimal actions are unique, offline RL achieves absolute zero sub-optimality error when K exceeds a (finite) instance-dependent threshold. To the best of our knowledge, these are the first 𝒪̃(1/K) bound and absolute zero sub-optimality bound respectively for offline RL with linear function approximation from adaptive data with partial coverage. We also provide instance-agnostic and instance-dependent information-theoretical lower bounds to complement our upper bounds.Thanh Nguyen-Tang, Ming Yin, Sunil Gupta, Svetha Venkatesh, Raman Arorawork_z5uyisimvvdjrhnq5wbdgc55xuWed, 23 Nov 2022 00:00:00 GMTSearch for electroweak production of supersymmetric particles in compressed mass spectra with the ATLAS detector at the LHC
https://scholar.archive.org/work/syrsys6kc5cv5ohkityim6kady
Two analyses searching for the production of supersymmetric particles through the electroweak interaction are presented: the chargino search, targeting the pair production of charginos decaying into W bosons and neutralinos, and the displaced track search, looking for charged tracks arising from the decays of higgsinos into pions. These searches target compressed phase spaces, where the mass difference between the next-to-lightest and lightest supersymmetric particle is relatively small. The searches use proton-proton collision data collected at a centre-of-mass energy of 13 TeV with the ATLAS detector at the LHC. In the chargino search, the targeted mass difference between charginos and neutralinos is close to the mass of the W boson. In such phase space, the chargino pair production is kinematically similar to the WW background, making the chargino signal experimentally challenging to be discriminated from the WW background. Machine learning techniques are adopted to separate the supersymmetric signal from the backgrounds. The results exclude chargino masses up to about 140 GeV for mass splittings down to about 100 GeV, superseding the previous results in particularly interesting regions where the chargino pair production could have hidden behind the looking-alike WW background. In the displaced track search, the mass difference between the produced sparticles and the lightest neutralinos goes down to 0.3 GeV. The experimental signature has a low momentum charged track with an origin displaced from the collision point. The results show that the analysis has the sensitivity to exclude different hypotheses for higgsino masses up to 175 GeV if no excess is observed in data. For lower masses, the larger signal cross-section allows to achieve higher significance for different mass splitting scenarios. All these signal hypotheses have not been probed by any existing analysis of LHC data.Eric Ballabenework_syrsys6kc5cv5ohkityim6kadyTue, 22 Nov 2022 00:00:00 GMT