Information Entropy's Discrimination on Gross Error with Origin
WU Xian-qiu · Physical Experiment of College · 2012
The non-statistical method which is base on the theory of information entropy can achieve gross error discrimination effectively when the data is not to much or the distribution is unclear.Basing on the example of linear fitting of magnetic field strength and excitation currency in the steady with Origin software,it shows that the method is viable and practical on removing outliers in physics experiments.