- Title: Empirical learning (housing)
- Goal: Given the data set, produce a learning system
that has the best out sample performance.
- Milestones:
- (1) Please go through the
housing training data and
description of data .
- (2) Using the data in any way you would like, hand us
- a learning machine that, you think, has the best
mean square test error (1/N * sum_i (f(xi)-g(xi))^2))
- and an estimate of what the test error will be.
- Further Issues:
- Tentative Schedule: week 1-4: milestone 1, 2;
- References:
- The optimization algorithms from
experiment 3.
- "Neural Networks for Pattern Recognition" by Chris Bishop.
- Caretaker: Zehra Cataltepe