Gradient descent finds global minima of deep neural networks
Gradient descent finds a global minimum in training deep neural networks despite the
objective function being non-convex. The current paper proves gradient descent achieves zero …
objective function being non-convex. The current paper proves gradient descent achieves zero …
Optimality and approximation with policy gradient methods in markov decision processes
Policy gradient (PG) methods are among the most effective methods in challenging reinforcement
learning problems with large state and/or action spaces. However, little is known about …
learning problems with large state and/or action spaces. However, little is known about …
The 2021 report of the Lancet Countdown on health and climate change: code red for a healthy future
The Lancet Countdown is an international collaboration that independently monitors the health
consequences of a changing climate. Publishing updated, new, and improved indicators …
consequences of a changing climate. Publishing updated, new, and improved indicators …
[HTML][HTML] The 2022 report of the Lancet Countdown on health and climate change: health at the mercy of fossil fuels
The 2022 report of the Lancet Countdown is published as the world confronts profound and
concurrent systemic shocks. Countries and health systems continue to contend with the …
concurrent systemic shocks. Countries and health systems continue to contend with the …
[HTML][HTML] Targeted gene disruption of matrix metalloproteinase-9 (gelatinase B) suppresses development of experimental abdominal aortic aneurysms
R Pyo, JK Lee, JM Shipley, JA Curci… - The Journal of …, 2000 - Am Soc Clin Investig
Abdominal aortic aneurysms represent a life-threatening condition characterized by chronic
inflammation, destructive remodeling of the extracellular matrix, and increased local …
inflammation, destructive remodeling of the extracellular matrix, and increased local …
Exact post-selection inference, with application to the lasso
We develop a general approach to valid inference after model selection. At the core of our
framework is a result that characterizes the distribution of a post-selection estimator …
framework is a result that characterizes the distribution of a post-selection estimator …
Microbiologically influenced corrosion
BJ Little, JS Lee - Oil and Gas Pipelines, 2015 - Wiley Online Library
The term microbiologically influenced corrosion (MIC) is used to designate corrosion due to
the presence and activities of microorganisms, that is, those organisms that cannot be seen …
the presence and activities of microorganisms, that is, those organisms that cannot be seen …
Green product purchase intention: Some insights from a developing country
This paper examines from the motivational perspective of the Theory of Reasoned Action (TRA)
how individual values and attitudes in a developing country influence purchase intention …
how individual values and attitudes in a developing country influence purchase intention …
Matrix completion has no spurious local minimum
Matrix completion is a basic machine learning problem that has wide applications, especially
in collaborative filtering and recommender systems. Simple non-convex optimization …
in collaborative filtering and recommender systems. Simple non-convex optimization …
[PDF][PDF] Gradient descent only converges to minimizers
We show that gradient descent converges to a local minimizer, almost surely with random
initial-ization. This is proved by applying the Stable Manifold Theorem from dynamical …
initial-ization. This is proved by applying the Stable Manifold Theorem from dynamical …