scatteR: Generating instance space based on scagnostics

TL;DR

scatteR is a novel data generation method that uses Scagnostics measures and an optimization process to create datasets with specific structural characteristics, offering an alternative to traditional model-based synthetic data methods.

Contribution

scatteR introduces a new approach to synthetic data generation by leveraging Scagnostics and optimization, focusing on data structure rather than model parameters.

Findings

Generates 50 data points in under 30 seconds

Achieves an average RMSE of 0.05 in matching target structures

Provides a pedagogical tool for teaching statistical data characteristics

Abstract

Traditional synthetic data generation methods rely on model-based approaches that tune the parameters of a model rather than focusing on the structure of the data itself. In contrast, Scagnostics is an exploratory graphical method that captures the structure of bivariate data using graph-theoretic measures. This paper presents a novel data generation method, scatteR, that uses Scagnostics measurements to control the characteristics of the generated dataset. By using an iterative Generalized Simulated Annealing optimizer, scatteR finds the optimal arrangement of data points that minimizes the distance between current and target Scagnostics measurements. The results demonstrate that scatteR can generate 50 data points in under 30 seconds with an average Root Mean Squared Error of 0.05, making it a useful pedagogical tool for teaching statistical methods. Overall, scatteR provides an entry point for generating datasets based on the characteristics of instance space, rather than relying on model-based simulations.