However, the difficulty of using and generating content is a complex factor that suggests difficulty when implementing more complicated models.
The success of the bifurcation experiment documenting model predicted dynamic changes by means of controlled and replicated laboratory experiments gives one the confidence to investigate other, more complicated model-predicted bifurcation sequences, including sequences involving chaotic attractors.
Multiple linear regression models, commonly used in environmental science, have a number of limitations: (1) all feature variables must be instantiated to obtain a prediction, and (2) the inclusion of categorical variables usually yields more complicated models.
Total speed-up is model-dependent and ranges from 5- to 100-fold, with larger, more complicated models generally having better relative performance (Fig. 5B).
Since the cost of additional parameters in AIC is only 2 k, the AIC favors more complicated models compared to BIC.
Furthermore, PAUP* provides a wide range of pairwise tree-distance measures, from simple absolute differences to more complicated model-based corrected ones.
Since the test set is completely independent of the parameter estimation process, selection will not be biased towards more complicated models.
This kind of information leads to high-dimensional and low-sample size (HDLSS) data sets (i.e., p ≫ n where p and n are the number of covariates and patients) which pose tremendous challenges to effective statistical inference especially for the time-to-event due to the presence of censoring and the use of much more complicated models.
For this reason, we have confidence that the proposed method is robust within ±10% noise ratio (more complicated models and trials on P2 can be found in part I of additional file 1). Figure 1 shows the responses of one trial.
The proposed procedure is based on the naïve Bayes approximation, therefore, the optimization of necessary parameters is performed only once and separately for each variable, thus resulting computationally fast, while later steps of the procedure enable to calculate more complicate models and choose the best one, without any further optimization.
