Results 131 to 140 of about 10,610,136 (340)

Efficient Neural Network Robustness Certification with General Activation Functions

, 2018
Finding minimum distortion of adversarial examples and thus certifying robustness in neural network classifiers for given data points is known to be a challenging problem.
Chen, Pin-Yu, Daniel, Luca, Hsieh, Cho-Jui, Weng, Tsui-Wei, Zhang, Huan   +4 more
core  

Tunable Mass Transport and Enhanced Electrochemical Performance of CO2 Laser Engraved Electrodes

Advanced Materials Technologies, EarlyView.
This work introduces vector‐mode CO2 laser engraving as a rapid strategy to fabricate electrochemical sensors with tunable mass transport. Spiral electrodes can behave as disk electrodes, outperforming conventional raster‐engraved devices in current density and sensitivity.
Davi M. De Farias, Gabriel N. Meloni, Thiago R. L. C. Paixão   +2 more
wiley   +1 more source

Accuracy of Dynamic Modulus Models of Asphalt Mixtures Containing Reclaimed Asphalt (RA)

Applied Sciences
The dynamic modulus (∣E*∣) is a fundamental mechanical parameter for studying the performance of hot mix asphalt and simulating its viscoelastic behaviour under different loading and thermal conditions.
Majda Belhaj, Jan Valentin, Nicola Baldo
doaj   +1 more source

Simultaneous optimization of shape, thickness and topology of shell structures using sigmoid function

Nihon Kikai Gakkai ronbunshu, 2019
Advancement of computer technologies as well as the developments of structural materials and construction methods have enabled us to design a so-called free-form shell, which has complex shape and topology that cannot be categorized to traditional shapes.
Shinnosuke FUJITA, Yoshihiro KANNO
doaj   +1 more source

Rotary 4D Printing of Programmable Metamaterials on Sustainable 4D Mandrel

Advanced Materials Technologies, EarlyView.
The paper introduces open, rotary 4D printing of programmable tubular metamaterials. It presents an open, non‐planar path‐to‐G‐code pipeline and shape‐changing rotary mandrel that enable support‐free fabrication and multi‐stimuli actuation. Experiments and FEA validate the mentioned features and stress redistribution.
Hesam Soleimanzadeh, Mahdi Bodaghi, Marzieh Jamalabadi, Bernard Rolfe, Ali Zolfagharian   +4 more
wiley   +1 more source

Empowering Random Walk Link Prediction Algorithms in Complex Networks by Adapted Structural Information

IEEE Access
In the link prediction problem a relevant algorithm running over a network attempts to determine whether a link between two nodes will exist in the future, given that it is not present at the moment.
Paraskevas Dimitriou, Vasileios Karyotis
doaj   +1 more source

A Comparison of Growth Curve Models for Forensically Important Flies Using Sigmoid Functions [PDF]

, 2020
Ji Eun Park, Su Jin Jeong, Sang Hyun Park, Sang Eon Shin, Seong Hwan Park, Tae‐Young Moon, Jae Won Lee   +6 more
openalex   +1 more source

Enhancing Convolutional Neural Network Performance through Optimized Sigmoid Activation Function Modeling [PDF]

, 2023
Y. Jin Youn, Jin-Keun Hong
openalex   +1 more source

Approximation order to a function in C(R) by superposition of a sigmoidal function

Applied Mathematics Letters, 2002
Abstract We investigate the approximation error to a continuous function defined on the whole real line by superposition of a sigmoidal function. With the minimal constraints on a continuous function, we show that the approximation order by 3n superposition of a sigmoidal function is O (1/n).
Nahmwoo Hahm, Bum Il Hong
openaire   +2 more sources

Generalizing the sigmoid function using continuous-valued logic

Fuzzy Sets Syst., 2022
J. Dombi, T. Jónás
semanticscholar   +1 more source