QUAYE, S. Vlastnosti Cauchyho rozdělení a jejich užití [online]. Brno: Vysoké učení technické v Brně. Fakulta strojního inženýrství. 2022.
During our regular meetings I revealed that Mr. Quaye does not have some basic knowledge in necessary parts of optimization, probability and statistics. However, by discussions and fulfillment of my assignments, he gained the knowledge and was able to apply it with my assistance in the problems connected with the topic of the thesis. On the other hand, Mr Quaye programmed the code for simulations of ROC curves in Python on his own. Unfortunately, after seeing the attachment of the submitted thesis, I found out that it (in some cases) meant copying a code for the empirical ROC curve from the internet (without referring to it). However, in my opinion, the copied code doesn’t calculate the empirical ROC curve as can be also seen in Figures 26-31. The student was warned about the possible incorrectness of the plotted estimate. The properties of the Cauchy distribution in Chapter 3 and robust parameter estimates in Chapter 4 can be found in various resources. Nevertheless, the Chapter 5 on ROC contains some novel results. The bi-Cauchy ROC is derived, its inflection points are found for some specific setting and various estimates of the bi-Cauchy ROC curve based on different robust estimates of the location parameters are compared in simulations visually and by AUC. Unfortunately, here Mr. Quaye derived the inflection points incorrectly after my last proofreading, probably by false calculations. The obtained coordinates of the inflection points of the ROC curves don’t lie in the interval . I also noticed that the reference to M. Pepe's book was not corrected after my warning. In conclusion, I would also like to mention that Mr. Quaye was meeting me regularly, he was trying to prepare for the meeting as much as possible, he was studying the resources and searching for others on his own and was always very polite.
The diploma thesis consists of three main parts. The first part deals with the description of the Cauchy distribution and methods of estimating its parameters. The second part deals with ROC curves and their estimation. The third part deals with Bi-Cauchy ROC curves, by generating them and from the generated parameter estimation curves. The work is written clearly without major typos. The pictures are clear and have a uniform shape. There are some incorrect wording and remarks, such as - on page 17 - bottom - impossibility to calculate "tails". - def. 3.1 - formula 4.15 It can therefore be stated that the student has fulfilled the assignment of the diploma thesis. I recommend diploma thesis to defend and I evaluate good.
eVSKP id 145975