Results 91 to 100 of about 186,399 (262)
Blockwise Self-Supervised Learning at Scale
Publisher Copyright: © 2024, Transactions on Machine Learning Research. All rights reserved.Current state-of-the-art deep networks are all powered by backpropagation.
Lecun, Yann +3 more
core
Advanced Engineering Materials, EarlyView.We develop a data‐driven method to derive the mathematical expressions of the Flory–Huggins interaction parameter χ for the swelling behavior of temperature–responsive hydrogels. Starting from initial assumptions of χ, our workflow combines Bayesian optimization, Flory–Rehner theory, and symbolic regression to generate candidate χ expressions.
Yawen Wang +2 more
wiley +1 more source
Putting the Science in Computer Science
, 2020 Modern science relies heavily on computers, and programming ability is moving from a useful skill to an indispensable prerequisite for scientific research. Both science and computer science classes can take advantage of this connection.
Illinois Mathematics and Science Academy
core
Microstructure Evolution of a VMnFeCoNi High‐Entropy Alloy After Synthesis, Swaging, and Annealing
Advanced Engineering Materials, EarlyView.The synthesis and processing (rotary swaging and annealing) of the novel VMnFeCoNi alloy is investigated, alongside the estimation of the grain size effect on hardness. Analysis of a wide grain size range of recrystallized microstructures (12–210 µm) reveals a low annealing twin density.
Aditya Srinivasan Tirunilai +6 more
wiley +1 more source
Development of an AutoML web application for machine learning tasks [PDF]
openAutomated Machine Learning, also known as AutoML, is a new and florid field of data science that aims at symplifying the process of designing, development and usage of typical machine learning models.
BELLO, ANDREA
core
Conditional gradient methods via stochastic path-integrated differential estimator
, 2021 We propose a class of novel variance-reduced stochastic conditional gradient methods. By adopting the recent stochastic path-integrated differential estimator technique (SPIDER) of Fang ct al.
Sra, Suvrit
core
A Lightweight Procedural Layer for Hybrid Experimental–Computational Workflows in Materials Science
Advanced Engineering Materials, EarlyView.We unveil a prototype hybrid‐workflow framework that fuses automatedcomputation with hands‐on experiments. Built atop pyiron, a lightweight, parameterized layer translates procedure descriptions into executable manual steps, syncing instrument settings, human interventions, and data capture in real‐time today.
Steffen Brinckmann +8 more
wiley +1 more source
Memory-Based Dual Gaussian Processes for Sequential Learning
, 2023 Sequential learning with Gaussian processes (GPs) is challenging when access to past data is limited, for example, in continual and active learning. In such cases, errors can accumulate over time due to inaccuracies in the posterior, hyperparameters, and
Chang, Paul E. +4 more
core
PASTA‐ELN: Simplifying Research Data Management for Experimental Materials Science
Advanced Engineering Materials, EarlyView.Research data management faces ongoing hurdles as many ELNs remain complex and restrictive. PASTA‐ELN offers an open‐source, cross‐platform solution that prioritizes simplicity, offline access, and user control. Its in tuitive folder structure, modular Python add‐ons, and open formats enable seamless documentation, FAIR data practices, and easy ...
S. Brinckmann, G. Winkens, R. Schwaiger
wiley +1 more source
Transfer in Reinforcement Learning via Shared Features
, 2012 We present a framework for transfer in reinforcement learning based on the idea that related tasks share some common features, and that transfer can be achieved via those shared features.
Scheidwasser, Ilya +2 more
core