2019, Vol. 17, No. 2
Selle kollektsiooni püsiv URIhttp://hdl.handle.net/10492/5393
Sirvi
Sirvi 2019, Vol. 17, No. 2 Märksõna "classification and regression tree" järgi
Nüüd näidatakse 1 - 1 1
Tulemused lehekülje kohta
Sorteerimisvalikud
Kirje Estonian dairy farms’ technical efficiency and factors predicting it(2019) Luik-Lindsaar, Helis; Põldaru, Reet; Roots, Jüri; Estonian University of Life Sciences. Institute of Economics and Social SciencesMilk production is a complex process whose efficiency depends directly on the inputoutput ratio and indirectly on the decisions made at farm and animal level. Decisions made about farm hygiene, dairy cows’ milk yield, cows’ age at first calving etc. affect farms’ efficiency. The aim of this study is to provide an understanding of the factors that affect dairy farms’ technical efficiency. A two-stage approach was used in this study, consisting of a data envelopment analysis (DEA) in the first stage, and classification and regression tree (CART) in the second stage. DEA determined technical efficiency scores (TE), and CART enabled to detect the main factors that influenced efficiency in dairy farms. The analysis studied at the Estonian national level FADN dataset and Estonian Livestock Performance Recording data. 147 Estonian dairy farms were included in this analysis, all of which are specialized in dairy production. DEA results demonstrated that more than half of the farms (55%) were operating efficiently or rather efficiently (TE ≥ 0.900). CART results revealed that the main variables determining efficiency are milk yield per cow's lifetime (kg day-1 ), feed costs (€ kg milk-1 ), and somatic cell count (SCC; 103 ml-1 ). Milk yield per cow's lifetime is a complicated factor as it is influenced by a lot of components (e.g. milk yield, number of lactations, age at first calving, and calving interval), but if it is known at farm level, it is also a useful variable for predicting efficiency. Feed costs per milk kg is an economic variable, i.e. lower costs are related with higher efficiency. Better hygiene (lower SCC) is also related with higher efficiency. The analysis showed that integrating farm accounts data, herd-level genetic information, and milk quality attributes enables to use more specific factors to explain the variation of TE between dairy farms.