Evaluation of Stream Data by Formal Concept Analysis
Following article presents practical usage of the Formal Concept Analysis (FCA) for the evaluation of stream data recorded during a technological process. The main aim of this paper is to show possibilities of using FCA to detect anomalies in the data. Our attitude is based on the fact that although during the production process a large amount of input data is obtained, the size of conceptual lattice is relatively small, and therefore, it is possible to work with it in real-time. The conceptual lattice represents a model of production process, and this model is based on historical production data. The input data stream contains measurements on the production line and it is applied on the model of the production process. The result of this activity is to identify anomalies in the incoming data and their relationship with faulty products, including disclosure of possible causes of errors and also to obtain a histogram of quality for manufactured products.
Keywordsstream data quality of production formal concept analysis data mining
Unable to display preview. Download preview PDF.
- 3.Franceschet, M.: The Role of Conference Publications in CS. Communications of the ACM 53(12) (December 2010)Google Scholar
- 5.Ganter, B., Stumme, G., Wille, R. (eds.): Formal Concept Analysis. LNCS (LNAI), vol. 3626. Springer, Heidelberg (2005)Google Scholar
- 6.Ganter, B., Wille, R.: Applied Lattice Theory: Formal Concept Analysis. In: Grätzer, G.A. (ed.) General Lattice Theory, pp. 592–606. Birkhäuser (1997)Google Scholar
- 9.Khosravi, A., Nahavandi, S., Creighton, D.C.: Predicting Amount of Saleable Products Using Neural Network Metamodels of Casthouses. In: ICARCV, pp. 2018–2023. IEEE (2010)Google Scholar
- 12.Montgomery, D.C.: Introduction to Statistical Quality Control. John Wiley & Sons, New York (1985)Google Scholar