Runs test, random number generators

Another simple variant of this method is to test the model on completely random data. Generate a random series of numbers for your figure of merit y, and then run your model. You should get only noise - if you get meaningful results (i.e., high R2 and q2 values) out of random data, then there is something seriously wrong with your model. 

In practice, all simulation models are stochastic models, i.e., both input and output variables are random variables. In a simulation run, only one specific constellation of possible random variables can be generated, and only the corresponding simulation results can be analyzed. In the present case, the actual time consumption of each individual activity is calculated from the input duration and the attributes of the activity, the tools, and the persons. This input duration disperses between freely definable limits, normally distributed around a predicted mean value. The determination of this variation is acquired with random numbers and ranges to 99 percent between freely definable limits of 10, 20, or 30 percent. The random numbers are between zero and one they were tested for autocorrelations smaller than 0.005 for a sample of 1000 random variables (mi,. .., tiiooo)- By means of the Box-Muller Method [855], the equally distributed random numbers were converted into random numbers (zi,. ..,ziooo) with a normal distribution (p = 0, o- = 1) ... 

An important characteristic of a stochastic optimization algorithm is that it always converges to the optimal value, regardless of its initial state. To test this characteristic, PSO and MPSO runs were repeated with 5 different random seeds 0, 50,000, 100,000, 150,000 and 200,000. To ensure repeatability of the experiments, the generator state was set to some fixed value each time the optimization executes to ensure that the same set of random numbers are generated. 

The primary performance measures of a ligand-binding assay are bias/trueness and precision. These measures along with the total error are then used to derive and evaluate several other performance characteristics such as sensitivity (LLOQ), dynamic range, and dilutional linearity. Estimation of the primary performance measures (bias, precision, and total error) requires relevant data to be generated from a number of independent runs (also termed as experiments or assay s). Within each run, a number of concentration levels of the analyte of interest are tested with two or more replicates at each level. The primary performance measures are estimated independently at each level of the analyte concentration. This is carried out within the framework of the analysis of variance (ANOVA) model with the experimental runs included as a random effect [23]. Additional terms such as analyst, instmment, etc., may be included in this model depending on the design of the experiment. This ANOVA model allows us to estimate the overall mean of the calculated concentrations and the relevant variance components such as the within-run variance and the between-run variance. 

---