Browsing by Subject "classification trees"

Now showing 1 - 2 of 2

Access status: Open Access ,
Machine learning methods for diagnosing the causes of die-casting defects
(Wydawnictwa AGH, 2023) Okuniewska, Alicja; Perzyk, Marcin; Kozłowski, Jacek
The research was focused on analyzing the causes of high-pressure die-casting defects, more specifically on casting leakage, which is considered perhaps the most important and common defect. The real data used for modelling was obtained from a high-pressure die-casting foundry that manufactures aluminum cylinder blocks for the world's leading automotive brands. This paper compares and summarizes the results of applying advanced modelling using artificial neural networks, regression trees, and support vector machines methods to select artificial neural networks as the most effective method to perform a multidimensional optimization of process parameters to diagnose the causes of die-casting defects and to indicate the future research scope in this area. The developed system enables the prediction of the level of defects in castings with satisfactory accuracy and is therefore a highly relevant reference for process engineers of high-pressure foundries. This article indicates exactly which process parameters significantly influence the formation of a defect in a casting.
Access status: Open Access ,
Modeling transaction prices of properties based on qualitative and quantitative features
(2010) Jasińska, Elżbieta; Preweda, Edward; Ruchel, Jan
The article presents a solution which allows to take in account the qualitative and quantitative features while modeling transaction prices of real estate apartments. The research material information concerned dwellings located within the city of Krakow, which have been sold in the period November 2008 - March 2009. Authors propose extension of the existing ways of testing the real estate market by multi--dimensional analysis, what will allow the comparison of impact of variables without assigning them numerical values. The solution is presented by the C&RT method (Classification and Regression Trees), which does not require scaling these attributes, as it can be described in a qualitative scale. This paper describes the optimal parameters of such models, thanks to which the creation of appropriate size tree is possible, that is a tree which allows the identification of rules which develop the property market in the selected districts. This proposal extends the existing research by taking into account the qualitative and quantitative features. It allows the introduction of an additional attribute, which is the location of a premises by a chosen street, which has been overlooked until now. Apart from defining the principles developing the general price of the real estate transaction, the features of the property have been lined up, showing the necessity of taking under consideration the address of the property.