Data Collaboration Analysis applied to Compound Datasets and the Introduction of Projection data to Non-IID settings

TL;DR

This paper introduces DCPd, an improved distributed machine learning method using projection data, which enhances prediction accuracy for compound datasets in non-IID settings, addressing federated learning limitations.

Contribution

The paper proposes DCPd, a novel data collaboration analysis method utilizing auxiliary PubChem data to improve non-IID compound data predictions.

Findings

DCPd outperforms FedAvg and DC in non-IID settings.

DCPd maintains high accuracy with minimal label bias impact.

Federated learning struggles with non-IID compound data, but DCPd mitigates this.

Abstract

Given the time and expense associated with bringing a drug to market, numerous studies have been conducted to predict the properties of compounds based on their structure using machine learning. Federated learning has been applied to compound datasets to increase their prediction accuracy while safeguarding potentially proprietary information. However, federated learning is encumbered by low accuracy in not identically and independently distributed (non-IID) settings, i.e., data partitioning has a large label bias, and is considered unsuitable for compound datasets, which tend to have large label bias. To address this limitation, we utilized an alternative method of distributed machine learning to chemical compound data from open sources, called data collaboration analysis (DC). We also proposed data collaboration analysis using projection data (DCPd), which is an improved method that utilizes auxiliary PubChem data. This improves the quality of individual user-side data transformations for the projection data for the creation of intermediate representations. The classification accuracy, i.e., area under the curve in the receiver operating characteristic curve (ROC-AUC) and AUC in the precision-recall curve (PR-AUC), of federated averaging (FedAvg), DC, and DCPd was compared for five compound datasets. We determined that the machine learning performance for non-IID settings was in the order of DCPd, DC, and FedAvg, although they were almost the same in identically and independently distributed (IID) settings. Moreover, the results showed that compared to other methods, DCPd exhibited a negligible decline in classification accuracy in experiments with different degrees of label bias. Thus, DCPd can address the low performance in non-IID settings, which is one of the challenges of federated learning.