Tutorials

Here are some tutorials to help you get started with Minerva and learn how to use it to train models for various tasks. The tutorials are written in Jupyter notebooks and can be run interactively. The tutorials are organized by topic.

Getting Started

In particular, we demonstrate how to use Minerva to train models for two distinct tasks using supervised learning with state-of-the-art architectures from the literature:

  • Human Activity Recognition (HAR): This task involves classifying a person’s activity based on time-series data collected from smartphone sensors, such as accelerometers and gyroscopes. It is formulated as a time-series classification problem.

  • Seismic Facies Classification: This task focuses on segmenting seismic images into different facies classes. It is approached as a semantic segmentation problem on 2D image data.

Seismic Facies Classification Models

Human Activity Recognition Models

Experiments