Student:	Cumhur ALEVLİ
Supervisor:	Yrd. Doç. Dr. İbrahim YILDIRIM
Department:	Orman Endüstri Müh.
Institution:	Graduate School of Natural and Applied Sciences
University:	Karadeniz Technical University Turkey

Title of the Thesis:	PRODUCTION PROJECTION AND FOREIGN TRADE ANALYSIS OF FOREST PRODUCTS INDUSTRY
Level:	M.Sc.
Acceptance Date:	3/6/2016
Number of Pages:	182
Registration Number:	i3051

Summary:

The position of forest products industry in the national and international trade is

very important. Forest products industry has back and forth connections with many sectors

and it can be an input for these sectors, in this respect, commercially importance of this

industry comes to the fore, too. When it is mentioned about forest products, a lot of

mixture of products come to mind. In this study, the forest products which are obtained

from FAOSTAT database and having useful data sets for the economic analysis examined

in detail. The study is divided into two parts, in first part the production and foreign trade

values of forest products examined in specific years using table and graphs and the impact

of the countries on these products is tried to show with the year 2014 values. Likewise the

EU (European Union) countries are also evaluated within themselves.

In the second part, based on the forest products produced between 1993-2014 years

in Turkey and Germany, it is aimed to estimate the production quantity of these products

till the year 2023. Multiple regression analysis and artificial neural networks method were

used for the estimation processes and it is determined that which method is preferable

mathematically. In both methods, the same independent variables used for estimation of a

product production. When choosing the independent variables, the model which gives the

best coefficient of determination was used. 10 products and 9 independent variables for

Turkey, 10 products and 7 independent variables for Germany were used to do estimating

processes.

Key Words: Forest products, Production, Multiple regression analysis, Artificial neural

networks, Turkey, Germany, European Union