Results 61 to 70 of about 22,732 (252)

On Scalable Inference with Stochastic Gradient Descent

CoRR, 2017
In many applications involving large dataset or online updating, stochastic gradient descent (SGD) provides a scalable way to compute parameter estimates and has gained increasing popularity due to its numerical convenience and memory efficiency. While the asymptotic properties of SGD-based estimators have been established decades ago, statistical ...
Yixin Fang, Jinfeng Xu, Lei Yang
openaire   +2 more sources

Ontology‐Aligned Structuring and Reuse of Multimodal Materials Data and Workflows Toward Automatic Reproduction

Advanced Engineering Materials, EarlyView.
Reproduction of stacking fault energy calculations from literature with a semi‐automated large language model‐assisted extraction procedure: extraction of simulation protocol, atomistic structures, computational parameters, and reported results, ontology alignment, knowledge graph construction and, finally, recomputation forvalidation.
Sepideh Baghaee Ravari, Abril Azocar Guzman, Sarath Menon, Stefan Sandfeld, Tilmann Hickel, Markus Stricker   +5 more
wiley   +1 more source

Inverse Identification of Energy‐Dependent Laser Absorptivity in NiTi Laser Powder‐Bed Fusion via Calibrated Melt Pool Simulation

Advanced Engineering Materials, EarlyView.
A combined experimental–computational framework identifies energy‐dependent laser absorptivity for NiTi in laser powder‐bed fusion, applicable to conduction and transition modes. Single‐track experiments and thermofluid smoothed particle hydrodynamics simulations are coupled through inverse analysis of melt pool geometry.
Mohamadreza Afrasiabi, Hadi Moutablaleh, Ignatius A. Setiawan, Mehrshad Mehrpouya   +3 more
wiley   +1 more source

New AI‐Assisted Approach for Expanding the Solution Space: Application to Lattice Structure Design

Advanced Engineering Materials, EarlyView.
This work introduces an innovative framework for designing structured materials by ex panding the design space through reparameterization of qualitative variables into continuous structural descriptors. Combined with machine‐learning‐based prediction and multi‐objective optimization, the approach enables the discovery of novel lattice architectures ...
G. H. Gahimbare, Y. Beideler, D. Chune, R. De Guio, T. Roland, H. Chibane   +5 more
wiley   +1 more source

Attentional-Biased Stochastic Gradient Descent

Trans. Mach. Learn. Res., 2020
In this paper, we present a simple yet effective provable method (named ABSGD) for addressing the data imbalance or label noise problem in deep learning. Our method is a simple modification to momentum SGD where we assign an individual importance weight to each sample in the mini-batch.
Qi Qi 0006, Yi Xu 0008, Wotao Yin, Rong Jin 0001, Tianbao Yang   +4 more
openaire   +3 more sources

Analysis of stochastic gradient descent in continuous time [PDF]

Statistics and Computing, 2021
AbstractStochastic gradient descent is an optimisation method that combines classical gradient descent with random subsampling within the target functional. In this work, we introduce the stochastic gradient process as a continuous-time representation of stochastic gradient descent.
openaire   +5 more sources

Multilayer Self‐Limiting Electrospray Deposition via Stepped Voltage Bias

Advanced Engineering Materials, EarlyView.
Self‐limiting electrospray deposition (SLED) uses a high voltage to generate and deposit a charged payload on a target surface. The coating retains its charge, repelling newly arriving material. SLED thickness can be decreased by applying a secondary bias to the target.
Madhuri Deb, S. Rahman Pejman, Alexander Jakimowicz, Jonathan P. Singer   +3 more
wiley   +1 more source

On the Generalization of Stochastic Gradient Descent with Momentum

J. Mach. Learn. Res., 2018
While momentum-based accelerated variants of stochastic gradient descent (SGD) are widely used when training machine learning models, there is little theoretical understanding on the generalization error of such methods. In this work, we first show that there exists a convex loss function for which the stability gap for multiple epochs of SGD with ...
Ramezani-Kebrya, Ali, Antonakopoulos, Kimon, Cevher, Volkan, Khisti, Ashish, Liang, Ben   +4 more
openaire   +4 more sources

All‐in‐One Analog AI Hardware: On‐Chip Training and Inference with Conductive‐Metal‐Oxide/HfOx ReRAM Devices

Advanced Functional Materials, EarlyView.
An all‐in‐one analog AI accelerator is presented, enabling on‐chip training, weight retention, and long‐term inference acceleration. It leverages a BEOL‐integrated CMO/HfOx ReRAM array with low‐voltage operation (<1.5 V), multi‐bit capability over 32 states, low programming noise (10 nS), and near‐ideal weight transfer.
Donato Francesco Falcone, Victoria Clerico, Wooseok Choi, Tommaso Stecconi, Folkert Horst, Laura Bégon‐Lours, Matteo Galetta, Antonio La Porta, Nikhil Garg, Fabien Alibart, Bert Jan Offrein, Valeria Bragaglia   +11 more
wiley   +1 more source

A new approach to training neural networks using natural gradient descent with momentum based on Dirichlet distributions

Компьютерная оптика, 2023
In this paper, we propose a natural gradient descent algorithm with momentum based on Dirichlet distributions to speed up the training of neural networks. This approach takes into account not only the direction of the gradients, but also the convexity of
R.I. Abdulkadirov, P.A. Lyakhov
doaj   +1 more source