www-ai.cs.tu-dortmund.de/de/LEHRE/PG/PG445/KDDcup2002/vogel_axelrod_2002.pdf
Microsoft Word - vogel.doc
section 3.1, we discuss the techniques actually used for our submission in the KDD Cup. In section 3.2, we discuss ideas for improvement after taking away the time constraints of the competition.
3.1 Decision [...] characters of the gene, and 3) A count of how many times a key word appeared within 60 characters of the gene. As it turned out, there was no general rule of which of the 3 were more predictive, as it [...] that could offer predictive values. High frequency gene aliases also were examined in this fashion.
2.3 Text Mining Variables Most of the key variables did not come from the data tables that were given. The …
