Numeric Encoding Options with Automunge

TL;DR

This paper explores the benefits of extended numeric feature encodings in deep learning for tabular data, using Automunge's transformation options to improve model performance through data augmentation.

Contribution

It introduces a comprehensive survey of numeric transformation options in Automunge and demonstrates a novel generalized noise injection method for data augmentation.

Findings

Extended encodings can enhance deep learning on tabular data.

Noise injection improves model performance with limited training data.

Family tree transformations enable flexible feature engineering.

Abstract

Mainstream practice in machine learning with tabular data may take for granted that any feature engineering beyond scaling for numeric sets is superfluous in context of deep neural networks. This paper will offer arguments for potential benefits of extended encodings of numeric streams in deep learning by way of a survey of options for numeric transformations as available in the Automunge open source python library platform for tabular data pipelines, where transformations may be applied to distinct columns in "family tree" sets with generations and branches of derivations. Automunge transformation options include normalization, binning, noise injection, derivatives, and more. The aggregation of these methods into family tree sets of transformations are demonstrated for use to present numeric features to machine learning in multiple configurations of varying information content, as may be applied to encode numeric sets of unknown interpretation. Experiments demonstrate the realization of a novel generalized solution to data augmentation by noise injection for tabular learning, as may materially benefit model performance in applications with underserved training data.