How to effectively adapt neural machine translation (NMT) models according to emerging cases without retraining? Despite the great success of neural machine translation, updating the deployed models online remains a challenge. Existing non-parametric
approaches that retrieve similar examples from a database to guide the translation process are promising but are prone to overfit the retrieved examples. However, non-parametric methods are prone to overfit the retrieved examples. In this work, we propose to learn Kernel-Smoothed Translation with Example Retrieval (KSTER), an effective approach to adapt neural machine translation models online. Experiments on domain adaptation and multi-domain machine translation datasets show that even without expensive retraining, KSTER is able to achieve improvement of 1.1 to 1.5 BLEU scores over the best existing online adaptation methods. The code and trained models are released at https://github.com/jiangqn/KSTER.
Linear regression methods impose strong constraints on regression models, especially on
the error terms where it assumes that it is independent and follows normal distribution, and
this may not be satisfied in many studies, leading to bias that can
not be ignored from the
actual model, which affects the credibility of the study.
We present in this paper the problem of estimating the regression function using the
Nadarya Watson kernel and k- nearest neighbor estimators as alternatives to the parametric
linear regression estimators through a simulation study on an imposed model, where we
conducted a comparative study between these methods using the statistical programming
language R in order to know the best of these estimations. Where the mean squares errors
(MSE) was used to determine the best estimate.
The results of the simulation study also indicate the effectiveness and efficiency of the
nonparametric in the representation of the regression function as compared to linear
regression estimators, and indicate the convergence of the performance of these two
estimates.
In this work , The fifth order non-polynomial spline functions is
used to solve linear volterra integral equations with weakly
singular kernel .
Numerical examples are presented to illustrate the applications
of this method and to compare the computed results with
other numerical methods.
In this research, we estimate the survival function by using three
nonparametric estimation methods and is to the method of Kaplan and
Meier, the method Gaussian Kernel and method Weibull Kernel.
Depending on the method of simulation and a complet
e real data from
reality. Where we proposed a new method to estimate the survival
function is to use Weibull Kernel and choose a new value to the optimal
bandwidth, which have an important role in the estimation process.
The aim of this paper is to discuss the necessary and sufficient conditions for the continuity of operator linear integral in Orlicz space on a compact set of functions realized with the terms of a lebegue measure of the Euclidean space ending dimens
ion and the use of the terms continuous measurement N-function definition continued N-function some theorems in Hilbert, Banach spaces. Then the research touched on the concept of the continued complementary N-function given, in order to discuss the terms of a continuing full for Integrative operator linear kernel which is studied, and to achieve qualities compact a functions set in W. Orlicz space and choose the best approximation for linear integrative operators. Finally a comparison is carried out between continuing full and weak convergence of the functional sequences in subspace of W. Orlicz space.
This study was carried out at three different forest sites in Syria in order to
determine the effect of changing rainfall, temperature and soil on kernel
productivity of stone pine (Pinus pinea L.)> these sites included: Jabal Alnabi
Mata, (Tartou
s province, L1), Dahr Alkhoser (Homs province, L2) and E′en
Jron site (Idleb province, L3). Results showed that kernel productivity of stone
pine per tree was 236.3, 252.8, 143 g per tree, and 177, 162.3, and 86.98 kg per
hectare in L1, L2, and L3, respectively. These differences were attributed due
to the variation in the composition, textured and fertility of the soil available in
the three locations studied. It was concluded that trees of stone pine grow better
and superior in Kernel productivity in humid and super-humid bioclimatic
zone.
This experiment was conducted at two ecologically different regions, Boka,
and Gellien, using 3 lines of X.triticosecale Wittmack (372, C.187, and C.G.2)
and 6 cultivars of wheat (5 of them were triticum durum Cham1, Cham3,
Cham5, Bohoth5, and Hau
rani, and one of triticum aestivum Cham6), to assess
the changes in water content and dry matter in the grains during the period
from anthesis to physiological maturity .The results showed that all genotypes
had the same moisture content curves, whereas it had seemed that the two
durum wheat cultivars (Cham1 and Bohoth5) exhibited a disturbance in the
moisture development curves in the first region, and the same observation was
noticed on (Cham1, Cham3, and Cham5) in the second region. However,
triticale lines had a higher test weight of 1000 grain compared with wheat
cultivars in the two regions, and there was a positive relationship between grain
fill duration and the weight of 1000 grain, whereas, there was a depression in
the test weight of wheat cultivars in the second region in comparison with the
first one, but it is associated with an increase of protein percentage, and this
might be attributed to temperature elevation during grain fill stage.
A landrace (Hourani-٢٧) and two newly introduced cultivars (ACSAD-٦٥
and Amra) were grown at five different locations (JUST, Shajarah, Hawarah,
Kharja and Turrah) with or without NP treatment.
