Verification of Economic and Agricultural Indicators With the Use of Statistical Methods on the Example of Individual Farms

Abstract

The paper describes basic methodology assumptions related to construction of Bayesian networks. The paper aims at preparation of data for modeling to obtain fresh knowledge on economic and agricultural database and will constitute the first stage of research. Variables (economic and agricultural indicators) with discreet values were used for analysis with the use of two step grouping technique and previous non-typical data were explored. The research was carried out on the group of three hundred individual fanns from Malopolskie and Swiçtokrzyskie Voiovedship. The knowledge obtained from analyses will be used in practice in agricultural engineering in order to support agricultural activity.

Application of Bayesian Networks for Forecasting Future Model of Farm

“…However, before the construction of the network was initiated, in the first stage of research, exploration of the input data including 300 individual farms from Małopolskie and Świętokrzyskie Voivodeship was carried out (Grotkiewicz et al, 2016). Exploration analysis aimed at possible searching for and exclusion of non-typical data.…”

“…In the first approach to build a model based on the Bayesian Networks all variables with discrete values were used after non-typical data were explored. Discrete values for variables (economic and agricultural indicators: PT, WP, WZ, PG, PC) were created as a result of analysis carried out in the first stage of research which uses a Two Step Cluster Analysis (Grotkiewicz et al, 2016).…”

“…The results of the research concerning data exploration and grouping of values with Two Step Cluster Analysis were presented in the paper by Grotkiewicz et al, (2016) which constitute at the same time the first stage of research.…”

“…Based on the previous analyses, concerning data preparation for modelling (Grotkiewicz et al, 2016), this paper constitutes the second stage of research, whose fundamental objective consists in an assessment of usefulness of the Bayesian Networks for forecasting development of economic and agricultural indicators with the use of various algorithms of Bayesian modelling and for obtaining new knowledge from the economic and agricultural data base. The effect of the research will consist in obtaining information on the probability distribution.…”

“…In the first approach to build a model based on the Bayesian networks all variables with discrete values present in the investigated data set were applied. Discrete values for varia-bles (economic and agricultural indicators) were formed as a result of the previous analysis which used Two Step Clustering technique (Grotkiewicz et al, 2016), where numbers and percentage values of discretized variables (numbers of clusters) referring to ranges of continuous variables values were presented in tables.…”