README.md

ml-finance

Projects

NLP on Earnings Release (10K, 8K)

Pata Processing (Week 0-2)

1. Download Notre Dame parsed 10K data--Week 1
2. Replicate JW paper (bags of words methond on 10K)--Week 1
3. Download and Parse SEC 8K data--Week 1-2
4. Stock Price database?

Algorithm Design (Week 3-8)

5. Potential directions
   5.1. Examine the impact on stock returns (i.e., market response) right after release
   5.2. Examine the potential usage on fraud dectection (see, PS paper)
   5.3. Examine the insider trading during 8-K trading gap
   http://blog.alphaarchitect.com/2015/10/21/insider-trading-8-k-trading-gap/#gs.7N=Q1wE
   5.4 For all the above direction: test advanced methods (e.g., Word2Vec, Doc2Vec, RNN, etc)

NLP on Twitter

Some background https://blog.twitter.com/en_us/topics/insights/2016/twitter-data-and-the-financial-markets.html http://ieeexplore.ieee.org/document/7069923/

Data Processing (Week 0-2) --Ty

1. Download twitter news of companies -- Week 1-2

Algorithm Design (Week 3-8)

2. Potential directions
   2.1. Examine the impact on stock returns (i.e., market response)
   2.2. Examine the potential usage on fraud dectection (applying the fog index, PS paper)
   2.3 For all the above direction: test advanced methods (e.g., Word2Vec, Paragraph2Vec, RNN, etc)

Plan

Week 0 (June 30):

1. Working on the code to replicate JW paper. --Mandy
2. Download 8K --David/Ty/Jake
3. Download twitter data --Ty

Week 1 (July 7):

1. Finished downloading the university of notre dame data. --Ty
2. Finished Replication of JW paper (as their dataset is different from notre dame data, data processing is still expected) --Mandy
3. Download Stock price return -- David
4. Continue downloading 8K data. -- Jake
5. Continue downloading twitter data. --Ty

July 7: Meet with Laura Li to discuss the direction.

Week 2 (July 14):

Week 3 (July 21):

Week 4 (July 28):

Week 5 (Aug 4):

Week 6 (Aug 11):

Week 7 (Aug 18):

Week 8 (Aug 25):