Design of effective in-silico adjusting method to support a doctor about the plan of administering medicine

摘要

We present the theory of the optimum running approximation of input signals using sample values of the corresponding output signals of multi-path analysis niters. The presented method uses time-limited running interpolation functions. As an application, we discuss design of in-silico smart adjusting-systems to support a doctor about the plan of administering medicine that is useful in personalized medical care. In this paper, firstly, we define a set of a finite number of signals in the initial set of signals and we present a one to one correspondence between each signal contained in the set of signals and the corresponding error of approximation of a certain finite time-interval. Secondly, based on this one-to-one correspondence, we prove that the presented running approximation minimizes various continuous worst-case measures of error at the same time. Certain reciprocal properties of the approximation are presented. Thirdly, extension to signal-estimation using multiple-input one-output system is presented. Finally, an application to the above in-silico adjusting method is discussed.