Detecting Patterns in North Korean Military Provocations: What Machine-learning Tells Us

Detecting Patterns in North Korean Military Provocations: What Machine-learning Tells Us Attached below are two csv files entitled, "peacearticles" and "threatarticles." We used the code below to write the xts/zoo object into a two .csv files. write.zoo(red, file = 'threatarticles.txt', row.names = TRUE) write.zoo(peace, file = 'peacearticles.txt', row.names = TRUE) Parameters for Cuts are as follows below: The Peace article cuts (dcuts), these 5 cuts are combined to create one object, from which we randomly select our desired ratio (7:10) 697 articles, from a database of 80,000 articles. The total database is about 89,000 articles. The Threat article cuts (ncuts) are 7 days prior to an incident. ## yeonpyeong naval skirmish ncut1=df.xts['1999-06-07/1999-06-14'] dcut1=df.xts['1997-01-01/1999-04-14'] ## North Korea vessels attack a South Korea naval ship ncut2=df.xts['2002-06-21/2002-06-28'] dcut2=df.xts['1999-08-14/2002-04-28'] # Naval vessels from North and South Korea exchange fire near the border ncut3 = df.xts['2009-11-02/2009-11-09'] dcut3=df.xts['2002-08-28/2009-09-09'] ## North Korea sinks the South Korean naval cheonan with a torpedo ncut4 = df.xts['2010-03-18/2010-03-25'] dcut4=df.xts['2010-05-25/2010-09-15'] ## North Korea fires artillery shells at South KOrea's Yeonpyeong island ncut5 = df.xts['2010-11-15/2010-11-22'] dcut5=df.xts['2011-01-22/2013-07-03']