01955nam a2200193 a 450000100080000000500110000800800410001910000190006024501260007926001900020552012320039565000170162765000170164465300160166165300140167770000230169170000260171470000210174020734672020-01-07       2017    bl uuuu     u00u1 u    #d1 aCAMARGO, L. de  aAnalysis of regional patterns of grain production and storage in Brazilba clustering application.h[electronic resource]  aIn: CONFERENCE INTERNATIONAL CONSORTIUM ON APPLIED BIOECONOMY RESEARCH, 21., 2017, Berkeley. Bioeconomy in transition: new tools and new actors. Berkeley: University of Californiac2017  aThe aim of this study is to identify patterns in the production and storage of grains in Brazil from 1980 to 2015 and analyze the conditions of similarity of each group. The technique of data mining Clustering was applied through the algorithm k-means to transform the historical series of grain storage and production into a group of similarities. The results show that the application of the mining technique coherently delimited the regions, which had a direct relation with the expansion of the agricultural frontier in Brazil, proving to be efficient for this application. Four clusters were formed. The first is composed by States of the South Region referring to consolidated crops since the 1960s. The second is composed of States of the Midwest and Southeast regions, revealing the second phase of the country's agricultural expansion. The third is made by states that do not have agricultural potential by soil and climate conditions, with an inexpressive agriculture. Finally, the fourth cluster shows the new Brazilian agricultural frontier called MATOPIBA (Maranhão, Tocantins, Piauí and Bahia), which has reached records in the production and export of grains driven by the installation of new transport routes.  aagribusiness  aAgronegócio  aData mining  aLogistics1 aOLIVEIRA, A. L. de1 aOLIVEIRA, S. R. de M.1 aCOLETI, J. de C.