2063 research outputs found
Sort by
Addressing Attribute Bias with Adversarial Support-Matching
Get PDFWhen trained on diverse labelled data, machine learning models have proven themselves to be a powerful tool in all facets of society. However, due to budget limitations, deliberate or non-deliberate censorship, and other problems during data collection, certain groups may be under-represented in the labelled training set. We investigate a scenario in which the absence of certain data is linked to the second level of a two-level hierarchy in the data. Inspired by the idea of protected attributes from algorithmic fairness, we consider generalised secondary "attributes" which subdivide the classes into smaller partitions. We refer to the partitions defined by the combination of an attribute and a class label, or leaf nodes in aforementioned hierarchy, as groups. To characterise the problem, we introduce the concept of classes with incomplete attribute support. The representational bias in the training set can give rise to spurious correlations between the classes and the attributes which cause standard classification models to generalise poorly to unseen groups. To overcome this bias, we make use of an additional, diverse but unlabelled dataset, called the deployment set, to learn a representation that is invariant to the attributes. This is done by adversarially matching the support of the training and deployment sets in representation space using a set discriminator operating on sets, or bags, of samples. In order to learn the desired invariance, it is paramount that the bags are balanced by class; this is easily achieved for the training set, but requires using semi-supervised clustering for the deployment set. We demonstrate the effectiveness of our method on several datasets and realisations of the problem
Thermally driven fluid convection in the incompressible limit regime
Get PDFWe consider a scaled Navier–Stokes–Fourier system describing the motion of a compressible, heat-conducting, viscous fluid driven by inhomogeneous boundary temperature distribution together with the gravitational force of a massive object placed outside the fluid. We
identify the limit system in the low Mach/low Froude number regime for the ill prepared initial
data. The fluid is confined to a bounded cavity with acoustically hard boundary enhancing
reflection of acoustic waves
Enhanced Schmidt-number criteria based on correlation trace norms
No full textThe Schmidt number represents the genuine entanglement dimension of a bipartite quantum state. We derive simple criteria for the Schmidt number of a density matrix in arbitrary local dimensions, given that certain symmetric measurements exist. They are based on the trace norm of correlations obtained from seminal families of quantum measurements, specifically symmetric informationally complete measurements and mutually unbiased bases. Our criteria are strictly stronger than both the well-known fidelity witness criterion and the computable cross-norm or realignment criterion
Fault detection and identification for control systems in floating offshore wind farms: A supervised Deep Learning methodology
No full textThis study employs a data-driven Fault Detection and Isolation (FDI) methodology in Floating Offshore Wind Turbine (FOWT) farms. The main objective of the work lies in classifying faults impacting the components of the control subsystems across multiple turbines. Unlike existing research, the emphasis here is placed specifically on identifying and classifying non-critical faults, which may result in suboptimal farm performance without necessitating a shutdown. From a methodological perspective, a Deep Neural Network has been designed to solve the classification problem by providing a probability vector, the most probable class indicator of the true state. One of the major contributions of this work lies in its applicability to FOWT farms instead of being confined to individual devices, facilitating a comprehensive performance assessment at the global farm level. The integration of this data-driven methodology with tolerant control strategies might enable early intervention, mitigating the impact of these faults and enhancing overall power generation efficiency. The target case study is a three-FOWT farm modeled in a Simulink environment, allowing for the simulation of operational behavior under diverse conditions and various faults affecting sensors and actuators. This work considers ten distinct fault classes, including the healthy condition, and three possible faults for each FOWT: pitch angle sensor, pitch angle actuator, and generator speed sensor. These frequent faults pose challenges to the optimal functioning of the control system managing the FOWTs. The outcomes highlight that the estimated probability of the healthy state serves as a robust indicator for detecting unknown faults. Results also demonstrate the adequate efficacy of the method in pinpointing the fault origin. However, we observe confusion between pitch sensor and actuator faults that require further investigation for comprehensive understanding.HAZITEK program (ERROTAID project (ZL-2022/00317)),
ELKARTEK program (TCRINI project (KK-2023-0029)
European Horizon (HE) with LIASON project (GA 101103698), FUTURAL project (101083958), and MsC Staff Exchange project RECHARGED (101086413).
Yerai Peña-Sanchez is funded by the European Union’s Horizon 2020 research and innovation program under
the Marie Sklodowska-Curie grant agreements N◦ 1010342
CsPbI3-PVDF Composite Based Multimode Hybrid Piezo-Triboelectric Nanogenerator: Self-Powered Moisture Monitoring System
No full textFor several decades, the development of potential flexible electronics, such as electronic skin, wearable technology, environmental monitoring systems, and the Internet of Things (IoT) network, has been emphasized. In this context, piezoelectric nanogenerator (PENG) and triboelectric nanogenerator (TENG) are highly regarded due to their simple design, high output performance, and cost-effectiveness. On a smaller scale, self-powered sensor research and development based on piezo-tribo hybrid nanogenerators have lately become more popular. When a material in the triboelectric nanogenerator is a piezoelectric material, these two distinct effects can be coupled. Herein, we developed a multi-mode hybrid piezo-triboelectric nanogenerator using CsPbI3-PVDF composite film. The addition of CsPbI3 to PVDF significantly enhances its electroactive phase and dielectric property, thereby enhancing its surface charge density. The optimized value of electroactive phase (FEA >90%) is observed with 5 wt. % CsPbI3 loading in PVDF. Moreover, piezoelectric CsPbI3-PVDF composite-based nanogenerator was fabricated in three modes viz. nanogenerator in contact separation mode (TECS), single electrode mode (TESE), sliding mode (TES) and the output performance of all the devices was investigated. The fabricated TECS, TESE, and TES reveal the peak output power 3.08, 1.29, and 0.15 mW at an external load of 5.6 Megaohm. Through analysis of the contact angle measurement and experimental quantification, the hydrophilicity of the composite film has been identified. The hydrophobicity and moisture absorption capacity of CsPbI3-PVDF film make it an attractive option for self-powered humidity monitoring. The TENGs effectively powered several low powered electronic devices with just a few human finger taps. This study offers a high-performance PTENG device that is reliant on ambient humidity, which is a helpful step towards creating a self-powered sensor.JDC2022-049793-I/MCIN/AEI/10.13039/501100011033
RES, QHS-2023-2-003
Correlation constraints and the Bloch geometry of two qubits
Get PDFWe present an inequality on the purity of a bipartite state depending solely on the length difference of the local Bloch vectors. For two qubits this inequality is tight for all marginal states and so extends the previously known solution for the two-qubit marginal problem. With this inequality we construct a three-dimensional Bloch model of the two-qubit quantum state space in terms of Bloch lengths, providing a pleasing visualization of this high-dimensional state space. This allows to characterize quantum states by a strongly reduced set of parameters and to investigate the interplay between local properties of the marginal systems and global properties encoded in the correlations
Characterization of rankings generated by pseudo-Boolean functions
Get PDFIn this paper we pursue the study of pseudo-Boolean functions as ranking generators. The objective of the work is to find new insights between the relation of the degree
of a pseudo-Boolean function and the rankings that can be generated by these insights. Based on a characterization theorem for pseudo-Boolean functions of degree
, several observations are made. First, we verify that pseudo-Boolean functions of degree
, where
is the search space dimension, cannot generate all the possible rankings of the solutions. Secondly, the sufficient condition for a ranking to be generated by a pseudo-Boolean function of dimension
is presented, and also the necessary condition is conjectured. Finally, we observe that the same argument is not sufficient to prove which ranking can be generated by pseudo-Boolean functions of degree
.This research has been partially supported by the Spanish Ministry of Science and Innovation [PID2022-137442NB-I00, PID2019-104933GB-I00 funded by MCIN/AEI/10.13039/501100011033] and BCAM Severo Ochoa accreditation, Spain [CEX2021-001142-S]; the Basque Government, Spain [BERC 2022–2025. IT1504-22, IT1494-22]; and the University of the Basque Country UPV/EHU [GIU20/054]. Open Access funding provided by University of Basque Country
Collision-free tool motion planning for 5-axis CNC machining with toroidal cutters
Get PDFCollision detection is a crucial part of CNC machining, however, many state-of-the-art algorithms test collisions as a post-process, after the path-planning stage, or use conservative approaches that result in low machining accuracy in the neighborhood of the cutter’s contact paths. We propose a fast collision detection test that does not require a costly construction of the configuration space nor high-resolution sampling of the cutter’s axis and uses the information of the neighboring points to efficiently prune away points of the axis that cannot cause collisions. The proposed collision detection test is incorporated directly as a part of the tool motion-planning stage, enabling design of highly-accurate motions of a toroidal cutting tool along free-form geometries. We validate our algorithm on a variety of benchmark surfaces, showing that our results provide high-quality approximations with provably non-colliding motions
Corrections to scaling in the 2D φ4 model: Monte Carlo results and some related problems
Get PDFMonte Carlo (MC) simulations have been performed to refine the estimation of
the correction-to-scaling exponent ω in the 2D φ
4 model, which belongs to one of
the most fundamental universality classes. If corrections have the form ∝ L
−ω, then
we find ω = 1.546(30) and ω = 1.509(14) as the best estimates. These are obtained
from the finite-size scaling of the susceptibility data in the range of linear lattice sizes
L ∈ [128, 2048] at the critical value of the Binder cumulant and from the scaling of
the corresponding pseudocritical couplings within L ∈ [64, 2048]. These values agree
with several other MC estimates at the assumption of the power-law corrections and
are comparable with the known results of the ϵ-expansion. In addition, we have tested
the consistency with the scaling corrections of the form ∝ L
−4/3
, ∝ L
−4/3
lnL and ∝
L
−4/3/ lnL, which might be expected from some considerations of the renormalization
group and Coulomb gas model. The latter option is consistent with our MC data. Our
MC results served as a basis for a critical reconsideration of some earlier theoretical
conjectures and scaling assumptions. In particular, we have corrected and refined
our previous analysis by grouping Feynman diagrams. The renewed analysis gives
ω ≈ 4 − d − 2η as some approximation for spatial dimensions d < 4, or ω ≈ 1.5 in two
dimensions
Atomistic understanding of ductile-to-brittle transition in single crystal Si and GaAs under nanoscratch
No full textEnsuring ductile removal in a grinding process is crucial for achieving the desired finish on a hard and brittle single crystal. This study provides new insights into the material removal processes in Si and GaAs single crystals, exploring their deformation behaviour using Berkovich and Conical tips to replicate contact from a fixed abrasive grit. Experimental observations are compared with Molecular Dynamic (MD) simulations to uncover the atomistic deformation mechanisms during the ductile-to-brittle transition (DBT). Notable plastic deformation and minimal cracking were consistently observed in Si, irrespective of the tips used. MD simulations supported this observation, revealing pronounced chip formation indicative of ductile material removal. The resistance to cracking in Si was attributed to amorphization induced by localized high contact stresses. In contrast, GaAs showed a propensity for cracking, with MD simulations revealing dislocation and slip band formation, and cracks emerging in the areas of substantial plastic deformation. These findings address phenomena not previously discernible in experimental studies due to the challenge of real-time observation. Moreover, the tip geometry was shown to significantly influence stress distribution, impacting deformation and crack formation in GaAs. This study also reveals limitations in predicting the critical depth for DBT in both Si and GaAs through the amended Bifano, Dow, and Scattergood (aBDS) models and MD simulation, offering nuanced insights into these challenges that have not been extensively explored. It was found that the experimental results exceeded predictions by an order of magnitude. These discrepancies underscore the aBDS model's disregard for essential material properties and tip geometry, while the disparities between MD simulation and experiment are primarily attributed to the inherent limitations in the simulated length scales and challenges in detecting initial subsurface cracks