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.
Increase of humidity in building envelopes often leads to the growth of mold, which is one of important factors for evaluation of Sick Building Syndrome. The estimation of mold contamination level in buildings based on electronic nose application is considered as fast and early detection technique, however interpretation of readouts is quite complicated, mostly because the signals obtained from sensor arrays are multidimensional. Moreover, there is no single optimal reference method used in practice. The idea of the presented approach is to group the readouts from sensor array into homogeneous sets of observations, which are characterized by the different level of mold contamination. The signals analyzed in the original 8-dimensional space are characterized by high variability depending on the conditions prevailing in the tested rooms, while the set of readouts may have a rather complicated shape (spherical-shaped clusters or convex clusters). In such a situation, the cluster analysis method based on density of signals could be applied. The most well-known density- based clustering method is DBSCAN (Density-Based Spatial Clustering of Applications with Noise). Unlike k-means or k-median, DBSCAN does not require the number of clusters as a parameter. Instead, it infers the number of clusters based on the data, and it can discover clusters of arbitrary shape (for comparison, k-means usually discovers spherical clusters). DBSCAN requires two parameters: ε (eps) and the minimum number of points required to form a dense region (minPts).
REFERENCES
