RECORD
Learned Feature Importance Scores for Automated Feature Engineering
Machine Learning · Submitted 6 Jun 2024
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim AutoMAN learns feature transform importance end-to-end, incorporating a dataset's task target directly into feature engineering, resulting in state-of-the-art performance with significantly lower latency compared to alternatives.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Feature engineering has demonstrated substantial utility for many machine learning workflows, such as in the small data regime or when distribution shifts are severe. Thus automating this capability can relieve much manual effort and improve model performance. Towards this, we propose AutoMAN, or Automated Mask-based Feature Engineering, an automated feature engineering framework that achieves high accuracy, low latency, and can be extended to heterogeneous and time-varying data. AutoMAN is based on effectively exploring the candidate transforms space, without explicitly manifesting transformed features. This is achieved by learning feature importance masks, which can be extended to support other modalities such as time series. AutoMAN learns feature transform importance end-to-end, incorporating a dataset's task target directly into feature engineering, resulting in state-of-the-art performance with significantly lower latency compared to alternatives.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Feature engineering has demonstrated substantial utility for many machine learning workflows, such as in the small data regime or when distribution shifts are severe. | background 1.00 | 0.06 | 0.03 | asserted 1.00 |
| s1 | Thus automating this capability can relieve much manual effort and improve model performance. | background 0.85 | 0.06 | 0.11 | asserted 0.72 |
| s2 | Towards this, we propose AutoMAN, or Automated Mask-based Feature Engineering, an automated feature engineering framework that achieves high accuracy, low latency, and can be extended to heterogeneous and time-varying data. | contribution 0.98 | 0.89 | 0.25 | asserted 1.00 |
| s3 | AutoMAN is based on effectively exploring the candidate transforms space, without explicitly manifesting transformed features. | method 0.99 | 0.82 | 0.06 | asserted 1.00 |
| s4 | This is achieved by learning feature importance masks, which can be extended to support other modalities such as time series. | method 0.98 | 0.83 | 0.09 | asserted 0.90 |
| s5 | AutoMAN learns feature transform importance end-to-end, incorporating a dataset's task target directly into feature engineering, resulting in state-of-the-art performance with significantly lower latency compared to alternatives. | result 0.69 | 0.93 | 0.76 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Feature engineering has demonstrated substantial utility for many machine learning workflows, such as in the small data regime or when distribution shifts are severe. Thus automating this capability can relieve much manual effort and improve model performance. Towards this, we propose AutoMAN, or Automated Mask-based Feature Engineering, an automated feature engineering framework that achieves high accuracy, low latency, and can be extended to heterogeneous and time-varying data. AutoMAN is based on effectively exploring the candidate transforms space, without explicitly manifesting transformed features. This is achieved by learning feature importance masks, which can be extended to support other modalities such as time series. AutoMAN learns feature transform importance end-to-end, incorporating a dataset's task target directly into feature engineering, resulting in state-of-the-art performance with significantly lower latency compared to alternatives.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites · 3 sentences without a confident label
Cites as background 3 papers
-
A variety of approaches have been explored towards automated feature engineering, including search-based algorithms that explore the feature space for optimal combinations [ref] , meta-learning approaches that leverage past feature engineering experiences [ref] , and deep learning models that automatically extract complex features from data [ref] .
SAFE: Scalable Automatic Feature Engineering Framework for Industrial Tasks · Introduction · 0.99
-
[ref] also employs reinforcement learning towards automated feature engineering, by developing a feature pre-evaluation model to improve sample and feature efficiency and warm-start policy training.
Toward Efficient Automated Feature Engineering · Related work · 0.88
-
On time series data, [ref] models multivariate time series as a cumulative sum of trajectory increments to engineer time series features at scale.
Feature Programming for Multivariate Time Series Prediction · Related work · 0.91
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.