This paper briefly presents two methods for regression analysis that are frequently met in recent literature dealing with modelling from data. These are Gaussian process models and support vector machines. An illustrative example with one dimensional data set is used to present some differences between models obtained with the used methods.
B.03 Paper at an international scientific conference
COBISS.SI-ID: 27155751Client-server environmental information presentation system designed and built by MEIS has been upgraded. It contains advanced data processing tools and air quality database. User interface of this client-server system has been upgraded to achieve best intuitive level. Tools for 2D presentations of ozone forecasts have been added for presentation of air pollution ground concentrations. In addition the possibility of additional settings to adjust the presentation legend for further analysis has been added.
F.16 Improvements to an existing information system/databases
Meteorological and air pollution quality database has been designed and built. Database contains measured half-hourly and hourly values from 13 automatic air quality measuring stations distributed across Slovenia and maintained by Slovenian Environmental Agency for time period of 2 years (2012 and 2013). Additionally it contains also half-hourly prognostic data from meteorological model WRF maintained by MEIS and prognostic air quality data from QualeAria air pollution modelling system for time period of 2 years (2012 and 2013). For all parameters also different statistical processing (8-hour moving daily, monthly and annual values) has been built according to decree of air quality.
F.16 Improvements to an existing information system/databases