%0 Generic %1 lam2018neural %A Lam, Hoang Thanh %A Minh, Tran Ngoc %A Sinn, Mathieu %A Buesser, Beat %A Wistuba, Martin %D 2018 %K automl button feature learning machine neural obm one solvatio %T Neural Feature Learning From Relational Database %U http://arxiv.org/abs/1801.05372 %X Feature engineering is one of the most important but most tedious tasks in data science. This work studies automation of feature learning from relational database. We first prove theoretically that finding the optimal features from relational data for predictive tasks is NP-hard. We propose an efficient rule-based approach based on heuristics and a deep neural network to automatically learn appropriate features from relational data. We benchmark our approaches in ensembles in past Kaggle competitions. Our new approach wins late medals and beats the state-of-the-art solutions with significant margins. To the best of our knowledge, this is the first time an automated data science system could win medals in Kaggle competitions with complex relational database.

@misc{lam2018neural, abstract = {Feature engineering is one of the most important but most tedious tasks in data science. This work studies automation of feature learning from relational database. We first prove theoretically that finding the optimal features from relational data for predictive tasks is NP-hard. We propose an efficient rule-based approach based on heuristics and a deep neural network to automatically learn appropriate features from relational data. We benchmark our approaches in ensembles in past Kaggle competitions. Our new approach wins late medals and beats the state-of-the-art solutions with significant margins. To the best of our knowledge, this is the first time an automated data science system could win medals in Kaggle competitions with complex relational database.}, added-at = {2019-01-09T09:48:16.000+0100}, author = {Lam, Hoang Thanh and Minh, Tran Ngoc and Sinn, Mathieu and Buesser, Beat and Wistuba, Martin}, biburl = {https://www.bibsonomy.org/bibtex/29a664a354f4d0a034efe6c59a303f54d/thoni}, description = {Neural Feature Learning From Relational Database}, interhash = {a17df48bf1b1fd823299f7518525c45d}, intrahash = {9a664a354f4d0a034efe6c59a303f54d}, keywords = {automl button feature learning machine neural obm one solvatio}, note = {cite arxiv:1801.05372}, timestamp = {2019-01-09T09:50:16.000+0100}, title = {Neural Feature Learning From Relational Database}, url = {http://arxiv.org/abs/1801.05372}, year = 2018 }