Results 21 to 30 of about 295,233 (270)
Optimising continuous microstructures: a comparison of gradient-based and stochastic methods [PDF]
, 2004 This work compares the use of a deterministic gradient based search with a stochastic genetic algorithm to optimise the geometry of a space frame structure.
Hanna, S., Haroun Mahdavi, S.
core
The Caenorhabditis elegans DPF‐3 and human DPP4 have tripeptidyl peptidase activity
FEBS Letters, EarlyView.The dipeptidyl peptidase IV (DPPIV) family comprises serine proteases classically defined by their ability to remove dipeptides from the N‐termini of substrates, a feature that gave the family its name. Here, we report the discovery of a previously unrecognized tripeptidyl peptidase activity in DPPIV family members from two different species.
Aditya Trivedi, Rajani Kanth Gudipati
wiley +1 more source
Adaptive Gradient Descent without Descent
, 2019 We present a strikingly simple proof that two rules are sufficient to automate gradient descent: 1) don't increase the stepsize too fast and 2) don't overstep the local curvature. No need for functional values, no line search, no information about the function except for the gradients.
Malitsky, Yura, Mishchenko, Konstantin
openaire +2 more sources
On the Finite Time Convergence of Cyclic Coordinate Descent Methods [PDF]
, 2010 Cyclic coordinate descent is a classic optimization method that has witnessed a resurgence of interest in machine learning. Reasons for this include its simplicity, speed and stability, as well as its competitive performance on $\ell_1$ regularized ...
Saha, Ankan, Tewari, Ambuj
core
Near-Optimal Straggler Mitigation for Distributed Gradient Methods
, 2017 Modern learning algorithms use gradient descent updates to train inferential models that best explain data. Scaling these approaches to massive data sizes requires proper distributed gradient descent schemes where distributed worker nodes compute partial
Avestimehr, A. Salman +3 more
core +1 more source
Molecular bases of circadian magnesium rhythms across eukaryotes
FEBS Letters, EarlyView.Circadian rhythms in intracellular [Mg2+] exist across eukaryotic kingdoms. Central roles for Mg2+ in metabolism suggest that Mg2+ rhythms could regulate daily cellular energy and metabolism. In this Perspective paper, we propose that ancestral prokaryotic transport proteins could be responsible for mediating Mg2+ rhythms and posit a feedback model ...
Helen K. Feord, Gerben van Ooijen
wiley +1 more source
Peptide‐based ligand antagonists block a Vibrio cholerae adhesin
FEBS Letters, EarlyView.The structure of a peptide‐binding domain of the Vibrio cholerae adhesin FrhA was solved by X‐ray crystallography, revealing how the inhibitory peptide AGYTD binds tightly at its Ca2+‐coordinated pocket. Structure‐guided design incorporating D‐amino acids enhanced binding affinity, providing a foundation for developing anti‐adhesion therapeutics ...
Mingyu Wang +9 more
wiley +1 more source
Learning to learn by gradient descent by gradient descent
, 2016 The move from hand-designed features to learned features in machine learning has been wildly successful. In spite of this, optimization algorithms are still designed by hand. In this paper we show how the design of an optimization algorithm can be cast as a learning problem, allowing the algorithm to learn to exploit structure in the problems of ...
Andrychowicz, Marcin +7 more
openaire +2 more sources
Adaptive hierarchical hyper-gradient descent
International Journal of Machine Learning and Cybernetics, 2022 AbstractAdaptive learning rate strategies can lead to faster convergence and better performance for deep learning models. There are some widely known human-designed adaptive optimizers such as Adam and RMSProp, gradient based adaptive methods such as hyper-descent and practical loss-based stepsize adaptation (L4), and meta learning approaches including
Renlong Jie +3 more
openaire +2 more sources
Real‐time assay of ribonucleotide reductase activity with a fluorescent RNA aptamer
FEBS Letters, EarlyView.Ribonucleotide reductases (RNR) synthesize DNA building blocks de novo, making them crucial in DNA replication and drug targeting. FLARE introduces the first single‐tube real‐time coupled RNR assay, which enables isothermal tracking of RNR activity at nanomolar enzyme levels and allows the reconstruction of allosteric regulatory patterns and rapid ...
Jacopo De Capitani +4 more
wiley +1 more source