The project is based on the analysis of the «2013 American Community Survey» dataset published on Kaggle and released under the public domain license (CC0).
The task is to implement from scratch a learning algorithm for regression with square loss (e.g., ridge regression). The label to be predicted must be selected among the following 5 attributes, removing the remaining 4 from the dataset:
PERNP (Person's earnings) PINCP (Person's income) WAGP (Wages or salary income past 12 months) HINCP (Household income) FINCP (Family income)