Unsupervised machine learning (ML) is in the forefront for exploring geological and geophysical data because of its simplicity and interpretability. It can analyze small (also big) and unlabeled data as opposed to supervised ML, which often requires massive and labeled data. Unsupervised ML discovers hidden patterns in the data that are not visible in the human eye. For example, principal component analysis (PCA) finds hidden patterns in the data by reducing dimensionality and increasing interpretability. Non-negative matrix factorization (NMF) finds hidden signatures in a dataset that well characterize the dataset. K-means clustering discovers underlying patterns in a dataset by grouping similar data together. Autoencoders are also unsupervised ML that leverage neural networks to learn patterns in datasets. In this session, we welcome abstracts, which utilize traditional (e.g., PCA, NMF, k-means clustering, autoencoders) or novel unsupervised ML to explore and mine geological and geophysical data.

• H - Hydrology
• NG - Nonlinear Geophysics

• GSA: Geological Society of America -
• JpGU: Japan Geoscience Union -
• SEG: Society of Exploration Geophysicists -
• WCRP: World Climate Research Programme -