Unsupervised Machine Learning Via Transfer Learning and k-Means Clustering to Classify Materials Image Data

Unsupervised machine learning offers significant opportunities for extracting knowledge from unlabeled datasets and for achieving maximum machine learning performance. This paper demonstrates how to construct, use, and evaluate a high-performance unsupervised machine learning system for classifying images in a popular microstructural dataset. The Northeastern University Steel Surface Defects Database includes micrographs of six different defects observed on hot-rolled steel in a format that is convenient for training and evaluating models for image classification. We use the VGG16 convolutional neural network pre-trained on the ImageNet dataset of natural images to extract feature representations for each micrograph. After applying principal component analysis to extract signal from the feature descriptors, we use k-means clustering to classify the images without needing labeled training data. The approach achieves 99.4% +/- 0.16% accuracy, and the resulting model can be used to classify new images without retraining. This approach demonstrates an improvement in both performance and utility compared to a previous study. A sensitivity analysis is conducted to better understand the influence of each step on the classification performance. The results provide insight toward applying unsupervised machine learning techniques to problems of interest in materials science.

Automated summary

Unsupervised machine learning is used to classify images in a microstructural dataset with high accuracy without the need for labeled training data. By extracting features and conducting clustering on the Northeastern University Steel Surface Defects Database, the approach outperforms previous studies in terms of both performance and utility. The results of a sensitivity analysis provide insights for applying unsupervised machine learning techniques to materials science problems of interest.