Linear regression analysis of indices describing Latvian wood processing industry
2013
Krumins, J., Latvia Univ. of Agriculture, Jelgava (Latvia) | Smits, I., Latvia Univ. of Agriculture, Jelgava (Latvia) | Dagis, S., Latvia Univ. of Agriculture, Jelgava (Latvia) | Dubrovskis, D., Latvia Univ. of Agriculture, Jelgava (Latvia) | Arhipova, I., Latvia Univ. of Agriculture, Jelgava (Latvia)
To forecast how different indices describing Latvian forest industry will develop in the future, the decision support program which is based on the research of supply chains and production processes as well as on systematic analysis of the whole industry should be developed. One of the most common types of analysis is modelling of processes. In this paper processes of forestry and wood processing in Latvia are modelled by using general approach for system modelling, and each process is described as an abstract system where only its input and output values were identified. Correlations of input and output value time series with a simple and multiple regression analysis method were analyzed. Regression analysis was created for 77 pairs of indices which theoretically could have significant correlations. Input and output factor linear regression analysis for set of processes Forestry and Wood processing shows that not always there is significant correlation between time series of chosen model factors. Some of correlations between time series of input and output data in the model of forest sector even showed controversial results. The amount of information about processes and their quantitative indicators in Latvian wood processing industry is not sufficient for development of precise simulation models. There is need to develop a list of criteria for missing information and carry out its collection process.
Показать больше [+] Меньше [-]