Please use this zoom link for the event.
Embeddings in machine learning provide a way to create a concise, lower-dimensional representation of complex, unstructured data. Embeddings are commonly employed in natural language processing to represent words or sentences and in computer vision for transfer learning. In this talk, we present work on creating embeddings of weather forecast grids (the High Resolution Rapid Refresh model) using autoencoders. The resulting embeddings turn out to have some nice properties, and you can take advantage of these properties to implement use cases like image search, interpolation, and clustering. In meteorology, these would be useful in identifying case studies, advection/nowcasting, and climatology.
The UW Data Science Seminar is an annual lecture series at the University of Washington that hosts scholars working across applied areas of data science, such as the sciences, engineering, humanities and arts along with methodological areas in data science, such as computer science, applied math and statistics. Our presenters come from all domain fields and include occasional external speakers from regional partners, governmental agencies and industry.
All seminars will be hosted virtually for the 2020-2021 academic year, and are free and open to the public.