Stock prediction using various regularization techniques on linear regression.
Predict % yield of next quarter
- Split / Dividends - Ramya
- Market Cap (Share price times total shares) - Vronsky
- Volume - Ramya
- P/E (Price to Earnings Ratio)[Market value per share / Earnings per Share (EPS)] - Parker
- SMP (Standard & Poor rating) - Shawn
- Field of Business (Ag, Tech, Finance) - Dan
- Size (Employees / Stockholder) - Chris
- State - Dan
- Number of financial quarters this company has existed before our start date - Dan
- Kurtosis - Ramya
- Companies must have been created before 2000 and in USA
- We will train on 30 companies
- We are running 30 DISTINCT regressions - one per company.
- We are using a Neural Network from the DeepLearn Toolbox. https://github.com/rasmusbergpalm/DeepLearnToolbox
- All company data will start on the year of the most recently created company
- Features are taken on a per quarter basis
Tue 11/18: Everyone is assigned a feature. Research it and present findings. Don't need data right now, just need explanation of how to use it.