CS240A Project 2: Mushrooms

← Back to Project 2 home

• Data

  • Dataset description in Delve
  • The data I used: I just turned the data into a cleartext file and made it comma-delimited rather than space-delimited.

• Main scripts (explained in the report)

  • load.sql
  • partition.sql*
  • nbc.sql
  • test.sql
  • boost.sql*

  * I used "~" to delimit line endings for these scripts to prevent db2 from becoming confused with ";", which is the end of a line in a block, but not of the whole block. These scripts must be run with "-d~"

• Other scripts I found useful

  • show.sql: show the current data in all relevant tables in the database.
  • store.sql: store the current data in all relevant tables in the database.
  • restore.sql: restore the data from a previous store.

• Results

  • output.txt: the output from running the five main scripts in db2 with the -v option.
  • result.txt: the output of running show.sql after the five major scripts, i.e., the contents of the relevant tables in the database.
  • result_neq_in.txt: contents of the database after using Booster A: boosting records IN misclassified.
  • result_neq_notin.txt: contents of the database after using Booster B: boosting records NOT IN misclassified.
  • result_eq_in.txt: contents of the database after using Booster C: boosting records IN correctly classified.
  • result_eq_notin.txt: contents of the database after using Booster D: boosting records NOT IN correctly classified.