4 ms·
Is it possible to use ML to improve on training returns? Similar to neural architecture search or circuit design optimization, can we use it to maximize perform
by datameta 3y ago
Is it possible to use ML to improve on training returns? Similar to neural architecture search or circuit design optimization, can we use it to maximize performance and accuracy through training set selection/optimization?
- woadwarrior01 3y agoData subset selection is an age old problem / sub-field in ML. Feature selection is a special case of this broader problem.