Probabilistic graphical models for semi-supervised traffic classification

Traffic classification using machine learning continues to be an active research area. The majority of work in this area uses off-the-shelf machine learning tools and treats them as black-box classifiers. This approach turns all the modelling complexity into a feature selection problem. In this paper, we build a problem-specific solution to the traffic classification problem by designing a custom probabilistic graphical model. Graphical models are a modular framework to design classifiers which incorporate domain-specific knowledge. More specifically, our solution introduces semi-supervised learning which means we learn from both labelled and unlabelled traffic flows. We show that our solution performs competitively compared to previous approaches while using less data and simpler features. Copyright © 2010 ACM.

Estado forestal del Pinus caribaea Morelet var. Hondurensis en el Pinal, Laguna de Perlas, RAAS

Tbe present study was carried out in the strip of land coast between Pearls Lagoon community and the Caribbean sea, in the nearness of Pinar lagoon, 25 km to the north of Bluefields. The geographical coordinates of the area are 12" 13' N and 83" 42' west. Tbe climate presents an annua! rninfall of 4 250 mm, an avera· ge temperature of 26 "C and a relative humidity of 89 %. The topography is plain and tbe elevation oscillates between O· 10 masl. The ma.in objective of the study was to evaluate the forest conditions in which forest of P. Caribaea var .. Jwndurei'ISis ealled the Pinal are found. The methodology employed consisted of gathering all information using air pllotographs and topograpbieal maps to defined the area and stands group (designed A, B, C, D, E, F and G). Two block and seven stands were found, in which invent.ory lines with sample plots of 500 m1 each one were employed to measure the indivíduals with DBH over 10 cm. The intensily it shows ís of !5% and was measured diameter, height, age and other. In tbis forest there where found seven stands with a total area covered with pine of 312.42 has. The average age is 23 years and estirnated total volume of 97.4 ms of wood. The density is of 60,61 trees hectare, with an average volume of 13.02 m3 /has and a commercial volume of 8.29 m' 1 has. Where found lndices of place 12 (stands G), 9 y 6. Besides, the 6.3.64 % of the trees has a satisfactory tendency of growlh (quality l and 2). lt can be concluded that the resource is in badly state due to intensive exploitations and annual f!fCS; is a young forest (23 years old), dotninated by índi viduals of small diameters (10-25 cm) and medium height (5-25 m); the available areas or with possibilities of being planting are: estimated in 468.64 has. This forest has a half annual increment of 1.4.3 cm/year, which is considered excellent and is c1assified with Clas P-III, FAO Classification by forests stratification.