The words contained in this file might help you see if this file matches what you are looking for:

...Chapter automating feature engineering in supervised learning udayan khurana ibm research introduction challenges performing terminology and problem denition afewsimple approaches hierarchical exploration of transformations transformation graph optimal traversal policy through reinforcement finding e ective features without model training to predict useful miscellenious other related work opportunities resources abstract the process predictive modeling requires extensive en gineering it often involves given space typically using mathematical functions with objective reducing error for a target however there is no well dened basis domain knowledge intuition most all lengthy trial human attention involved overseeing this signi cantly inuences cost generation moreover when data presented not described labeled manual becomes an even more prohibitive task we discuss ways algorithmically tackle engineer ing context fe representation plays important role ectiveness algorithm instance figure d...