Miles Cranmer - Effective Use of Machine Learning in Astrophysics
Duration: 60 mins
Share this media item:
Embed this media item:
Embed this media item:
About this item
| Description: | (No description) |
|---|
| Created: | 2023-11-28 15:33 | ||
|---|---|---|---|
| Collection: | New Frontiers in Astrophysics: A KICC Perspective | ||
| Publisher: | University of Cambridge | ||
| Copyright: | Miles Cranmer | ||
| Language: | eng (English) | ||
| Distribution: |
World
(not downloadable)
|
||
| Keywords: | Astrophysics; Machine Learning; | ||
| Credits: |
|
||
| Explicit content: | No | ||
| Aspect Ratio: | 4:3 | ||
| Screencast: | No | ||
| Bumper: | UCS Default | ||
| Trailer: | UCS Default | ||
| Abstract: | The field of machine learning (ML) offers a powerful set of frameworks for addressing complex problems in astrophysics, ranging from emulating expensive simulations to performing anomaly detection in large datasets. This talk explores a diverse range of ML applications within astrophysics, highlighting the role of these methods in extracting insights from multidimensional and multimodal datasets. I will also discuss the major challenges of ML, such as model robustness, interpretability, uncertainty estimation, and incorporation of physical priors. In all, this presentation will provide astronomers with a pragmatic overview of machine learning’s capabilities and limitations, and how these techniques will continue to shape astrophysical discovery. |
|---|---|
Available Formats
| Format | Quality | Bitrate | Size | |||
|---|---|---|---|---|---|---|
| MPEG-4 Video | 640x360 | 1.25 Mbits/sec | 566.60 MB | View | ||
| WebM | 640x360 | 423.98 kbits/sec | 189.43 MB | View | ||
| iPod Video | 480x360 | 502.96 kbits/sec | 221.03 MB | View | ||
| MP3 | 44100 Hz | 249.95 kbits/sec | 111.68 MB | Listen | ||
| Auto * | (Allows browser to choose a format it supports) | |||||