Zgadzam się
Nasza strona zapisuje niewielkie pliki tekstowe, nazywane ciasteczkami (ang. cookies) na Twoim urządzeniu w celu lepszego dostosowania treści oraz dla celów statystycznych. Możesz wyłączyć możliwość ich zapisu, zmieniając ustawienia Twojej przeglądarki. Korzystanie z naszej strony bez zmiany ustawień oznacza zgodę na przechowywanie cookies w Twoim urządzeniu.
The mold infestation of buildings classified by Kohonen Self-Organizing Maps with boundaries determined by Ward clustering using multidimensional data from gas sensors
Mold infestation of buildings occurs when the moisture content of partitions increases,
and is a significant problem in building operation. This problem is substantial in terms of
architecture and building construction, residents’ health and aesthetic reasons. There are
numerous methods of evaluating mold infestation, among them important ones include
traditional biological, molecular microbiological, and chemical techniques. One of the newer
methods is application of gas sensors arrays, which form an electronic nose when combined with
a properly chosen data analysis algorithm. The critical issue connected with correct functioning
of an electronic nose is selection of the appropriate mathematical model enabling interpretation
and visualization of the results – multidimensional signals originating from sensors array. In this
work a Kohonen Self-Organizing-Map with hexagonal topology was used for presenting the
similarity between measurements of buildings that are in different stages of mold infestation, as
well as reference sample of clean air and decayed timber. On the two-dimensional visualization
of Kohonen map, the boundaries created by applying the hierarchical Ward clustering method
were superimposed. This procedure allowed showing which observation would be assigned to
which clusters connected with level of mold infestation.
