Data acquisition in modeling using neural networks and decision trees
Journal Title: Archiwum Odlewnictwa - Year 2011, Vol 11, Issue 2
Abstract
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.
Authors and Affiliations
R. Sika, Z. Ignaszak
Keywords
Related Articles
- EP ID EP72526
- DOI -
- Views 101
- Downloads 0
Effect of overheating degree on the solidification parameters of AlSi17CuNiMg silumin
Melting technology, especially the temperature of overheating and pouring above the liquidus point Tliq as well as the cooling rate, determine alloy macro - and microstructure, and hence also casting performance proper...
Criterion for selection the optimal physical and chemical properties of cobalt aluminate powder used in investment casting process
The aim of this work was to determine physical and chemical properties of cobalt aluminate (CoAl2O4) modifiers produced by different companies and the influence of different types of modifiers on the grain size of high t...
Assessment Criteria of Bentonite Binding Properties
The criteria, with which one should be guided at the assessment of the binding properties of bentonites used for moulding sands, are proposed in the paper. Apart from the standard parameter which is the active bentonite...
The comparative studies of ADI versus Hadfield cast steel wear resistance
The results of comparative studies of wear resistance of ADI versus high manganese Hadfield cast steel are presented. For evaluation of wear resistance three type of ADI were chosen. Two of them were of moderate strength...
Production of Steel Casts in Two-Layer Moulds with Alkaline Binders Part 1. Backing sand with the alkaline inorganic binder RUDAL
Steel casts in Z.N. POMET were produced in moulds made of the moulding sand Floster. This sand did not have good knocking out properties, required a significant binder addition (4.5-5.0 parts by weight), and the casting...