arXiv subjects

Ye Wei

Publications and source records attributed to Ye Wei.

25 records · Page 2Linked to original sources

arXiv2023

Traffic4cast at NeurIPS 2022 -- Predict Dynamics along Graph Edges from Sparse Node Data: Whole City Traffic and ETA from Stationary Vehicle Detectors

The global trends of urbanization and increased personal mobility force us to rethink the way we live and use urban space. The Traffic4cast competition series tackles this problem in a data-driven way, advancing the latest methods in machine learning for modeling complex spatial systems over time. In this edition, our dynamic road graph data combine information from road maps, $10^{12}$ probe data points, and stationary vehicle detectors in three cities over the span of two years. While stationary vehicle detectors are the most accurate way to capture traffic volume, they are only available in few locations. Traffic4cast 2022 explores models that have the ability to generalize loosely related temporal vertex data on just a few nodes to predict dynamic future traffic states on the edges of the entire road graph. In the core challenge, participants are invited to predict the likelihoods of three congestion classes derived from the speed levels in the GPS data for the entire road graph in three cities 15 min into the future. We only provide vehicle count data from spatially sparse stationary vehicle detectors in these three cities as model input for this task. The data are aggregated in 15 min time bins for one hour prior to the prediction time. For the extended challenge, participants are tasked to predict the average travel times on super-segments 15 min into the future - super-segments are longer sequences of road segments in the graph. The competition results provide an important advance in the prediction of complex city-wide traffic states just from publicly available sparse vehicle data and without the need for large amounts of real-time floating vehicle data.

cs.LG↗

arXiv2022

A mechanically strong and ductile soft magnet with extremely low coercivity

Soft magnetic materials (SMMs) serve in electrical applications and sustainable energy supply, allowing magnetic flux variation in response to changes in applied magnetic field, at low energy loss1. The electrification of transport, households and manufacturing leads to an increase in energy consumption due to hysteresis losses2. Therefore, minimizing coercivity, which scales these losses, is crucial3. Yet, meeting this target alone is not enough: SMMs in electrical engines must withstand severe mechanical loads, i.e., the alloys need high strength and ductility4. This is a fundamental design challenge, as most methods that enhance strength introduce stress fields that can pin magnetic domains, thus increasing coercivity and hysteretic losses5. Here, we introduce an approach to overcome this dilemma. We have designed a Fe-Co-Ni-Ta-Al multicomponent alloy with ferromagnetic matrix and paramagnetic coherent nanoparticles (~91 nm size, ~55% volume fraction). They impede dislocation motion, enhancing strength and ductility. Their small size, low coherency and small magnetostatic energy create an interaction volume below the magnetic domain wall width, leading to minimal domain wall pinning, thus maintaining the soft magnetic properties. The alloy has a tensile strength of 1336 MPa at 54% tensile elongation, extremely low coercivity of 78 A/m (<1 Oe), moderate saturation magnetization of 100 Am2/kg, and high electrical resistivity of 103 μΩ u Ohm cm.

cond-mat.mtrl-sci↗

arXiv2022

Machine learning-enabled high-entropy alloy discovery

High-entropy alloys are solid solutions of multiple principal elements, capable of reaching composition and feature regimes inaccessible for dilute materials. Discovering those with valuable properties, however, relies on serendipity, as thermodynamic alloy design rules alone often fail in high-dimensional composition spaces. Here, we propose an active-learning strategy to accelerate the design of novel high-entropy Invar alloys in a practically infinite compositional space, based on very sparse data. Our approach works as a closed-loop, integrating machine learning with density-functional theory, thermodynamic calculations, and experiments. After processing and characterizing 17 new alloys (out of millions of possible compositions), we identified 2 high-entropy Invar alloys with extremely low thermal expansion coefficients around 2*10-6 K-1 at 300 K. Our study thus opens a new pathway for the fast and automated discovery of high-entropy alloys with optimal thermal, magnetic and electrical properties.

cond-mat.mtrl-sci↗

arXiv2021

Revealing in-plane grain boundary composition features through machine learning from atom probe tomography data

Grain boundaries (GBs) are planar lattice defects that govern the properties of many types of polycrystalline materials. Hence, their structures have been investigated in great detail. However, much less is known about their chemical features, owing to the experimental difficulties to probe these features at the atomic length scale inside bulk material specimens. Atom probe tomography (APT) is a tool capable of accomplishing this task, with an ability to quantify chemical characteristics at near-atomic scale. Using APT data sets, we present here a machine-learning-based approach for the automated quantification of chemical features of GBs. We trained a convolutional neural network (CNN) using twenty thousand synthesized images of grain interiors, GBs, or triple junctions. Such a trained CNN automatically detects the locations of GBs from APT data. Those GBs are then subjected to compositional mapping and analysis, including revealing their in-plane chemical decoration patterns. We applied this approach to experimentally obtained APT data sets pertaining to three case studies, namely, Ni-P, Pt-Au, and Al-Zn-Mg-Cu alloys. In the first case, we extracted GB-specific segregation features as a function of misorientation and coincidence site lattice character. Secondly, we revealed interfacial excesses and in-plane chemical features that could not have been found by standard compositional analyses. Lastly, we tracked the temporal evolution of chemical decoration from early-stage solute GB segregation in the dilute limit to interfacial phase separation, characterized by the evolution of complex composition patterns. This machine-learning-based approach provides quantitative, unbiased, and automated access to GB chemical analyses, serving as an enabling tool for new discoveries related to interface thermodynamics, kinetics, and the associated chemistry-structure-property relations.

cond-mat.mtrl-sci↗

arXiv2021

Convolutional neural network-assisted recognition of nanoscale L12 ordered structures in face-centred cubic alloys

Nanoscale L12-type ordered structures are widely used in face-centred cubic (FCC) alloys to exploit their hardening capacity and thereby improve mechanical properties. These fine-scale particles are typically fully coherent with matrix with the same atomic configuration disregarding chemical species, which makes them challenging to be characterized. Spatial distribution maps (SDMs) are used to probe local order by interrogating the three-dimensional (3D) distribution of atoms within reconstructed atom probe tomography (APT) data. However, it is almost impossible to manually analyse the complete point cloud ($>10$ million) in search for the partial crystallographic information retained within the data. Here, we proposed an intelligent L12-ordered structure recognition method based on convolutional neural networks (CNNs). The SDMs of a simulated L12-ordered structure and the FCC matrix were firstly generated. These simulated images combined with a small amount of experimental data were used to train a CNN-based L12-ordered structure recognition model. Finally, the approach was successfully applied to reveal the 3D distribution of L12-type $δ^\prime$-Al3(LiMg) nanoparticles with an average radius of 2.54 nm in a FCC Al-Li-Mg system. The minimum radius of detectable nanodomain is even down to 5 Å. The proposed CNN-APT method is promising to be extended to recognize other nanoscale ordered structures and even more-challenging short-range ordered phenomena in the near future.

cond-mat.mtrl-sci↗

arXiv2020

Privacy-Preserving Distributed Projection LMS for Linear Multitask Networks

We develop a privacy-preserving distributed projection least mean squares (LMS) strategy over linear multitask networks, where agents' local parameters of interest or tasks are linearly related. Each agent is interested in not only improving its local inference performance via in-network cooperation with neighboring agents, but also protecting its own individual task against privacy leakage. In our proposed strategy, at each time instant, each agent sends a noisy estimate, which is its local intermediate estimate corrupted by a zero-mean additive noise, to its neighboring agents. We derive a sufficient condition to determine the amount of noise to add to each agent's intermediate estimate to achieve an optimal trade-off between the network mean-square-deviation and an inference privacy constraint. We propose a distributed and adaptive strategy to compute the additive noise powers, and study the mean and mean-square behaviors and privacy-preserving performance of the proposed strategy. Simulation results demonstrate that our strategy is able to balance the trade-off between estimation accuracy and privacy preservation.

cs.MA↗

arXiv2020

Machine-learning-enhanced time-of-flight mass spectrometry analysis

Mass spectrometry is a widespread approach to work out what are the constituents of a material. Atoms and molecules are removed from the material and collected, and subsequently, a critical step is to infer their correct identities based from patterns formed in their mass-to-charge ratios and relative isotopic abundances. However, this identification step still mainly relies on individual user's expertise, making its standardization challenging, and hindering efficient data processing. Here, we introduce an approach that leverages modern machine learning technique to identify peak patterns in time-of-flight mass spectra within microseconds, outperforming human users without loss of accuracy. Our approach is cross-validated on mass spectra generated from different time-of-flight mass spectrometry(ToF-MS) techniques, offering the ToF-MS community an open-source, intelligent mass spectra analysis.

cond-mat.mtrl-sci↗