cs6601p3

8 Commits 1 Branch 0 Tags

Author	SHA1	Message	Date
Woody Folsom	1898c35571	Updated build for Java 8. Version 6 is no longer supported as of Java SE JDK 25.	2026-02-10 11:15:09 -05:00
Woody Folsom	5270359b10	Removed dangerous code to populate the database - this must be retrieved from git in order to repopulate the headlines table. (This should not need to happen). PricePoller and ValidationSetCreator generate the 1, 2, 3-grams.txt and validation.txt files, respectively. MySQLHeadlineDaoImplTest reshuffles the training, validation datasets in 60-40 ratio.	2012-04-22 21:24:01 -04:00
Woody Folsom	6e3680426e	Added script to pull historical stock data and resulting data files (1 per company). Added code to generate average price change per 1, 2 and 3-gram. Added code to output average price change per headline for VALIDATION dataset.	2012-04-20 21:22:54 -04:00
Woody Folsom	eec32b19c1	Added ability to generate 1, 2, 3-gram models on a company/date-range basis using <UNK> to represent the initial appearance of a previously unknown word.	2012-04-16 14:03:16 -04:00
Woody Folsom	027adff2dd	HeadlinePuller can import all news articles for Fortune 50 within a date range from Yahoo Finance historical data, inserting the headline text into the headlines table of database cs6601p3 on woodyfolsom.net:3306. Limited to 25 headline per day (with many reptitions) per Yahoo Finance REST functionality.	2012-04-15 10:18:02 -04:00
Woody Folsom	d700d97124	Functional data import from Yahoo Finance news using YQL (Yahoo Query Language) and XPATH. Data is stuffed into MySQL database cs6601 on woodyfolsom.net.	2012-04-07 18:59:39 -04:00
Woody Folsom	a46e790059	Skeleton of Project proposal document in LaTeX and some useful research papers on HMMs, CRFs and news headline classification.	2012-04-01 20:19:26 -04:00
Woody Folsom	225efee2b7	Test commit.	2012-04-01 20:06:00 -04:00