On the Role of Glonass for the Development of the Russian Geodetic Reference Network

Abstract

Re ev vn ni iv vy yk kh h S Se er rg ge ey y a a , , T Ta at te ev vy ya an n S Su ur ri iy ya a b b a a C Ce en nt tr ra al l R Re es se ea ar rc ch h I In ns st ti it tu ut te e o of f t th he e F Fe ed de er ra al l S Sp pa ac ce e A Ag ge en nc cy y, , 4 4, , P Pi io on ne er rs sk ka ay ya a s st tr r. .1 14 41 10 07 70 0, , K Ko or ro ol le ev v ((M Mo os sc co ow w)), , R RF F.. Sergey.

Machine Learning–Based Prediction of Clinical Outcomes for Children During Emergency Department Triage

“…Similarly, in the random forest models, we used out-of-bag (left-out samples after bagging) estimation to measure the prediction errors . In the deep neural networks, to minimize potential overfitting, we used dropout that randomly removes portions of units in the network, ridge regularization that shrinks large coefficients, and batch normalization that normalizes the means and variances of layer inputs …”