ExploreKit: Automatic Feature Generation and Selection

Katz, Gilad; Shin, Eui Chul; Song, Dawn

doi:10.1109/icdm.2016.0123

Cited by 136 publications

(104 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…not hosted by SPARQL endpoints). Finally, it would be interesting to investigate techniques for automatic feature selection [22], such as those described in [9], where a novel machine learning-based feature selection method is used for predicting candidates features usefulness.…”

Section: Resultsmentioning

confidence: 99%

How Linked Data can Aid Machine Learning-Based Tasks

Mountantonakis¹,

Tzitzikas²

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The discovery of useful data for a given problem is of primary importance since data scientists usually spend a lot of time for discovering, collecting and preparing data before using them for various reasons, e.g., for applying or testing machine learning algorithms. In this paper we propose a general method for discovering, creating and selecting, in an easy way, valuable features describing a set of entities for leveraging them in a machine learning context. We demonstrate the feasibility of this approach by introducing a tool (research prototype), called LODsyndesisML, which is based on Linked Data technologies, that a) discovers automatically datasets where the entities of interest occur, b) shows to the user a big number of useful features for these entities, and c) creates automatically the selected features by sending SPARQL queries. We evaluate this approach by exploiting data from several sources, including British National Library, for creating datasets in order to predict whether a book or a movie is popular or non-popular. Our evaluation contains a 5-fold cross validation and we introduce comparative results for a number of different features and models. The evaluation showed that the additional features did improve the accuracy of prediction.

show abstract

Section: Resultsmentioning

confidence: 99%

How Linked Data can Aid Machine Learning-Based Tasks

Mountantonakis¹,

Tzitzikas²

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…In this section, we state motivation of our work: why we consider high-order cross features and why existing works do not suit our purpose. While most early works of automatic feature generation focus on second-order interactions of original features [5,6,19,21,34], trends have appeared to consider higher-order (i.e., with order higher than two) interactions to make data more informative and discriminative [2,26,33,42]. High-order cross features, just like other high-order interactions, can further improve the quality of data and increase predictive power of learning algorithms.…”

Section: Motivationmentioning

confidence: 99%

“…On the one hand, search-based feature generation methods employ explicit search strategies to construct useful features or feature sets. Many such methods focus on numerical features [11,20,21,35,36], and do not generate cross features. As for existing feature crossing methods [5,34], they are not designed, and are hence inefficient, to perform high-order feature crossing.…”

Section: Motivationmentioning

confidence: 99%

“…The performance of most existing automatic feature generation methods greatly depends on some hyper-parameters. Examples are the thresholds in ExploreKit [21], subsampling ratio in [5], and network architectures in deep-learning-based methods [7,14,26,33,42]. These hyper-parameters should be properly determined or carefully tuned by the user for each specific task.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

AutoCross

Luo¹,

Wang²,

Zhou³

et al. 2019

Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

Feature crossing captures interactions among categorical features and is useful to enhance learning from tabular data in real-world businesses. In this paper, we present AutoCross, an automatic feature crossing tool provided by 4Paradigm to its customers, ranging from banks, hospitals, to Internet corporations. By performing beam search in a tree-structured space, AutoCross enables efficient generation of high-order cross features, which is not yet visited by existing works. Additionally, we propose successive mini-batch gradient descent and multi-granularity discretization to further improve efficiency and effectiveness, while ensuring simplicity so that no machine learning expertise or tedious hyper-parameter tuning is required. Furthermore, the algorithms are designed to reduce the computational, transmitting, and storage costs involved in distributed computing. Experimental results on both benchmark and real-world business datasets demonstrate the effectiveness and efficiency of AutoCross. It is shown that AutoCross can significantly enhance the performance of both linear and deep models. 1

show abstract

“…Most of existing methods of automatic FE either generate a large set of possible features by predefined transformation operators followed by feature selection [3,7,15] or apply simple supervised learning (simple algorithm and/or simple meta-features derived from FE process) to recommend a potentially useful feature [4,5,9]. The former makes the process computationally expensive, which is even worse for complex features, while the latter significantly limits the performance boost.…”

Section: Introductionmentioning

confidence: 99%

Cross-data Automatic Feature Engineering via Meta-learning and Reinforcement Learning

Zhang

Hao

Fogelman-Soulié³

2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Feature Engineering (FE) is one of the most beneficial, yet most difficult and time-consuming tasks of machine learning projects, and requires strong expert knowledge. It is thus significant to design generalized ways to perform FE. The primary difficulties arise from the multiform information to consider, the potentially infinite number of possible features and the high computational cost of feature generation and evaluation. We present a framework called Cross-data Automatic Feature Engineering Machine (CAFEM), which formalizes the FE problem as an optimization problem over a Feature Transformation Graph (FTG). CAFEM contains two components: a FE learner (FeL) that learns finegrained FE strategies on one single dataset by Double Deep Q-learning (DDQN) and a Cross-data Component (CdC) that speeds up FE learning on an unseen dataset by the generalized FE policies learned by Meta-Learning on a collection of datasets. We compare the performance of FeL with several existing state-of-the-art automatic FE techniques on a large collection of datasets. It shows that FeL outperforms existing approaches and is robust on the selection of learning algorithms. Further experiments also show that CdC can not only speed up FE learning but also increase learning performance.

show abstract

ExploreKit: Automatic Feature Generation and Selection

Cited by 136 publications

References 8 publications

How Linked Data can Aid Machine Learning-Based Tasks

How Linked Data can Aid Machine Learning-Based Tasks

AutoCross

Cross-data Automatic Feature Engineering via Meta-learning and Reinforcement Learning

Contact Info

Product

Resources

About