Foto del docente

Elisabetta Carfagna

Full Professor

Department of Statistical Sciences "Paolo Fortunati"

Academic discipline: SECS-S/01 Statistics

Research

Keywords: sequential sampling land use statistics adaptive sampling remote sensing data area and multiple frames sample designs spatial statistics geographic information systems (GIS)

SPECIALISATION / RESEARCH INTERESTS

· Sampling methods, area, list and multiple frames sample designs; use of administrative data for agricultural statistics;

· Use of remote sensing, geographic information systems and global positioning systems for agricultural and rural statistics;

· Adaptive sampling, design of experiments, and other statistical methods for continuous innovation, quality control and impact evaluation;

· Design and analysis of agricultural, rural, land use, forestry and agro-environmental statistics;

· Analysis and use of geographic and geo-referenced information, impact of spatial resolution, change of support and transformations of spatial data, spatial statistics;

· Analysis and forecasting of prices of cereals;

· Design and analysis of national and international statistics for agricultural and rural development;

· Statistical methods for analysing financial markets;

· Rural building and land use analysis and planning;

· Agro-meteorological yield forecasting models.

Main research subjects concern: sequential adaptive methods, area sample surveys, incomplete and multiple frames, use of remote sensing data for statistical purposes, spatial statistics, also developing relationships between the various research fields. Main applications of these methods are the quality control and the validation of geo-referenced data bases; the optimization of area and point frame sample designs, a model for the study of changes in the rural built environment, the production of agricultural land cover statistics and the analysis of the book market and of reading habits.

A research subject is the development of efficient adaptive sequential methods which, under general conditions, in the finite population sampling approach, guaranty unbiased estimates of the population parameters. Main applications of these methods are the quality control and the validation of geo-referenced data bases, particularly the photo-interpretation of remote sensing data, and the survey of land cover data in the presence of measurement errors.

Sample surveys, multiple frames sample designs, spatial statistics, statistical aspects of remote sensing data, analysis of survey data and of national and international statistics, with particular reference to agricultural, forestry, land use, environmental and agro-environmental data.

The relationship created between area frame sample designs and spatial statistics has the aim of optimizing area and point sample surveys through strategies which involve the spatial autocorrelation in the sample design and use auxiliary variables with geographic coordinates.

The link between area frame sample designs, the use of remote sensing data and adaptive sequential sampling has allowed to create a model for the studyof changes in the rural built environment.

A research field is the use of remote sensing data for producing agriculture and land use statistics, also evaluating the cost effectiveness and the combined use of various information sources for creating monitoring systems of the agricultural environments.

Other research subjects are inferential methods for estimating distribution functions, regressive models for qualitative and quantitative data applied to territorial and medical problems and the analysis of the book market and of reading habits.