Statistical analysis of continuous datasets showed that concentrations of N and P changed over the 21 year period.
Numerical experiments on various benchmark datasets showed that the proposed algorithm is fast and gives improved generalization performance over the existing methods.
Comparison of the derived regression equation with five literature datasets showed that the model can predict mean flow velocities in overland flow at a reasonable accuracy as long as the mean velocity is below 0.4 m s−1.
Application of the model to diverse C and N datasets showed that incorporating latency into the model of C mineralization, and using C to drive N dynamics, allows heterogeneous data from many different soil amendments to be described by the same model.
Furthermore, the joined analysis of the different datasets showed that small differences in soil moisture dynamics, especially at the upper and lower bounds of soil moisture (at maximum porosity and wilting point of the soils) can have a large influence on the soil moisture patterns and their autocorrelation structure.
Additional in silico analysis on simulated MIRU-VNTR genotype population datasets showed that the MIRU-VNTR pattern similarity could partly arise due to probabilistic increase of acquiring homoplasy among subpopulations, thus the similar MIRU-VNTR profiles of M. avium strains even in close geographical proximity should be interpreted with caution.
Experimental results on a variety of datasets showed that our method outperforms conventional methods and both steps in our system are essential for the overall performance.
Experimental results on three datasets showed that the new method improved the robustness of feature, in comparison to baseline systems trained on the same speech datasets.
However, analysis of the breast-height five-ring and the whole-disc datasets showed that approximately 45%% of the variation in these datasets was due to differences between sites.
Experiments using new datasets showed that the prediction made by ANN had an excellent overall performance and was very well correlated to the 3-month ischemic lesion on the T2-weighted image.
Two stimulated datasets showed that the reduced bias results in an increased range and reduced false positive of the estimate, but lower correlation with the true distribution due to increased variance.
