Diff2Dist: Learning Spectrally Distinct Edge Functions, with Applications to Cell Morphology Analysis

TL;DR

This paper introduces Diff2Dist, a neural network-based method to learn spectrally descriptive edge weights for graphs, improving graph discrimination in biological image analysis and simulation comparison.

Contribution

It generalizes the Graph Diffusion Distance to be differentiable and trainable with neural networks for enhanced graph classification and biological applications.

Findings

Contrastive training improves graph discrimination.

Learned distances enhance k-NN classifier performance.

Application to biological graphs reveals meaningful similarities.

Abstract

We present a method for learning "spectrally descriptive" edge weights for graphs. We generalize a previously known distance measure on graphs (Graph Diffusion Distance), thereby allowing it to be tuned to minimize an arbitrary loss function. Because all steps involved in calculating this modified GDD are differentiable, we demonstrate that it is possible for a small neural network model to learn edge weights which minimize loss. GDD alone does not effectively discriminate between graphs constructed from shoot apical meristem images of wild-type vs. mutant \emph{Arabidopsis thaliana} specimens. However, training edge weights and kernel parameters with contrastive loss produces a learned distance metric with large margins between these graph categories. We demonstrate this by showing improved performance of a simple k-nearest-neighbors classifier on the learned distance matrix. We also demonstrate a further application of this method to biological image analysis: once trained, we use our model to compute the distance between the biological graphs and a set of graphs output by a cell division simulator. This allows us to identify simulation parameter regimes which are similar to each class of graph in our original dataset.