Data acquisition in modeling using neural networks and decision trees

• Autorzy: Sika R. , Ignaszak Z.
• Języki publikacji: EN

Abstrakty

EN

The paper presents a comparison of selected models from area of artificial neural networks and decision trees in relation with actual conditions of foundry processes. The work contains short descriptions of used algorithms, their destination and method of data preparation, which is a domain of work of Data Mining systems. First part concerns data acquisition realized in selected iron foundry, indicating problems to solve in aspect of casting process modeling. Second part is a comparison of selected algorithms: a decision tree and artificial neural network, that is CART (Classification And Regression Trees) and BP (Backpropagation) in MLP (Multilayer Perceptron) networks algorithms. Aim of the paper is to show an aspect of selecting data for modeling, cleaning it and reducing, for example due to too strong correlation between some of recorded process parameters. Also, it has been shown what results can be obtained using two different approaches: first when modeling using available commercial software, for example Statistica, second when modeling step by step using Excel spreadsheet basing on the same algorithm, like BP-MLP. Discrepancy of results obtained from these two approaches originates from a priori made assumptions. Mentioned earlier Statistica universal software package, when used without awareness of relations of technological parameters, i.e. without user having experience in foundry and without scheduling ranks of particular parameters basing on acquisition, can not give credible basis to predict the quality of the castings. Also, a decisive influence of data acquisition method has been clearly indicated, the acquisition should be conducted according to repetitive measurement and control procedures. This paper is based on about 250 records of actual data, for one assortment for 6 month period, where only 12 data sets were complete (including two that were used for validation of neural network) and useful for creating a model. It is definitely too small portion in case of artificial neural networks, but it shows a scale of danger of unprofessional data acquisition.

Słowa kluczowe

EN

computer aiding; foundry proces; data acquisition; data mining; decision tree; neural network

PL

wspomaganie komputerowe; proces odlewniczy; gromadzenie danych; eksploracja danych; drzewo decyzyjne; sieć komputerowa

• Wydawca: Komisja Odlewnictwa Polskiej Akademii Nauk Oddział w Katowicach
• Czasopismo: Archives of Foundry Engineering
• Rocznik: 2011
• Tom: Vol. 11, iss. 2
• Strony: 113--122
• Opis fizyczny: Bibliogr. 17 poz., rys., tab., wykr.

Twórcy

autor

Sika R.

autor

Ignaszak Z.

• Division of Foundry, Poznań University of Technology, Piotrowo 3, 61-138 Poznań, Poland,

Bibliografia

• [1] Ignaszak Z., Virtual Prototyping in foundry. Poznan University of Technology, Poznań 2002 (in Polish).
• [2] Ignaszak Z., The specification and examples of on-line validation methods needed for quality forecasting systems for industrial castings [in:] Innovations in castiing part III, Edited by J. Sobczak, Institut of Casting, Cracow 2009 (in Polish).
• [3] Perzyk M., Kochański A., The Possibility of artificial neural networks application for casting processes modeling", Soldification of Metals and Alloys, No. 38, 1998 (in Polish).
• [4] Larose T., Knowledge discovering from data. Introduction to Data Mining, PWN, Warsaw 2006 (in Polish).
• [5] Łapczyński M., Classification trees in customer satisfaction and loyalty studies. Statsoft 2003 [w:] http://www.statsoft.pl/czytelnia/marketing/drzewa.pdf (in Polish).
• [6] Sika R., Ignaszak Z., Acquisition and preliminary preparation of non-homogenous data needed to Data Mining systems on the example of foundry industry, Archives of Production Engineering and Automation, Poznan University of Technology, Poznan 2009 (in Polish).
• [7] Sika R., Ignaszak Z., Data Mining in the foundry industry - problems recording and collection of non-homogenous data. Conference Modeling of Casting and Foundry Processes - Śrem 2008 (in Polish).
• [8] Sika R., Ignaszak Z., Data Analysis - system to optmalization quality of production processes in foundry. User guide. Poznań - Leszno 2009 (in Polish, non published).
• [9] Sika R., Ignaszak Z., After implementation KonMas-final program - use to cast production process analyze - W6 department - Foundry in Śrem, International Symposium - Modeling of casting and foundry processes, Poznań - Śrem, October 2006.
• [10] Sika R., Study on the SAP R/3 system structure and possibility of adapting to the management and quality control in Śrem Foundry, Master's thesis under the direction of Z. Ignaszak, Poznan University of Technology, 2006 (in Polish).
• [11] Ignaszak Z., Sika R., Exploration system for selected production data and its testing in the foundry, Archives of Production Engineering and Automation, Poznan University of Technology, Posen 2008 (in Polish).
• [12] Ignaszak Z., Sika R., Application of Data Mining systems for compilation of data in virtualization systems explored in the design and manufacturing process control in found, Unpublished work, Poznan University of Technology, 2009 (in Polish).
• [13] Perzyk M., Data mining in the foundry. The potential, problems and projects. Presentation at the XI International Symposium on Modeling of Casting and Foundry Processes, October 26-27, 2008, Poznan-Srem (Poland).
• [14] Wyrozumski T., How do I make the data were clean? Proceedings PLOUG IX Conference, October 2003, Kościelisko Poland (in Polish).
• [15] R. Sika, Z. Ignaszak, Quality Assurance in the foundry industry. Acquisition and preliminary development of heterogeneous, Archives of Mechanical Engineering and Automation, 2009.
• [16] Jakubski J., Dobosz, M., Application of neuronal networks for quality control of moulding sand. XXXIII Conference of National Foundryman Day, Faculty of Foundry Engineering Science and Technology, Krakow, 2009.
• [17] Perzyk M., Soroczyński A., Comparison of selected tools for generation of knowledge for foundry production, Archives of Foundry Engineering, Vol.8, Issue 4/2008.