Spectroscopy Approaches for Food Safety Applications: Improving Data Efficiency Using Active Learning and Semi-Supervised Learning

TL;DR

This paper investigates spectroscopy-based food safety analysis using machine learning, demonstrating that active learning and semi-supervised learning significantly reduce the need for labeled data, thereby improving data efficiency.

Contribution

The study introduces and compares active learning, semi-supervised learning, and their hybrid for spectroscopy data annotation in food safety applications, enhancing data efficiency.

Findings

AL reduces labeled data by 50%

SSL reduces labeled data by 25%

Hybrid approach improves data annotation efficiency

Abstract

The past decade witnesses a rapid development in the measurement and monitoring technologies for food science. Among these technologies, spectroscopy has been widely used for the analysis of food quality, safety, and nutritional properties. Due to the complexity of food systems and the lack of comprehensive predictive models, rapid and simple measurements to predict complex properties in food systems are largely missing. Machine Learning (ML) has shown great potential to improve classification and prediction of these properties. However, the barriers to collect large datasets for ML applications still persists. In this paper, we explore different approaches of data annotation and model training to improve data efficiency for ML applications. Specifically, we leverage Active Learning (AL) and Semi-Supervised Learning (SSL) and investigate four approaches: baseline passive learning, AL, SSL, and a hybrid of AL and SSL. To evaluate these approaches, we collect two spectroscopy datasets: predicting plasma dosage and detecting foodborne pathogen. Our experimental results show that, compared to the de facto passive learning approach, AL and SSL methods reduce the number of labeled samples by 50% and 25% for each ML application, respectively.