Category theory and functional programming for scalable statistical modelling and computational inference
45 mins 41 secs,
83.57 MB,
MP3
44100 Hz,
249.76 kbits/sec
Share this media item:
Embed this media item:
Embed this media item:
About this item
Description: |
Wilkinson, D
Tuesday 4th July 2017 - 16:15 to 17:00 |
---|
Created: | 2017-07-21 14:22 |
---|---|
Collection: | Scalable inference; statistical, algorithmic, computational aspects |
Publisher: | Isaac Newton Institute |
Copyright: | Wilkinson, D |
Language: | eng (English) |
Distribution: | World (downloadable) |
Explicit content: | No |
Aspect Ratio: | 16:9 |
Screencast: | No |
Bumper: | UCS Default |
Trailer: | UCS Default |
Abstract: | This talk considers both the theoretical and computational requirements for scalable statistical modelling and computation. It will be argued that programming languages typically used for statistical computing do not naturally scale, and that functional programming languages by contrast are ideally suited to the development of scalable statistical algorithms. The mathematical subject of category theory provides the necessary theoretical underpinnings for rigorous analysis and reasoning about functional algorithms, their correctness, and their scalability. Used in conjunction with other tools from theoretical computer science, such as recursion schemes, these approaches narrow the gap between statistical theory and computational implementation, providing numerous benefits, not least automatic parallelisation and distribution of algorithms. |
---|
Available Formats
Format | Quality | Bitrate | Size | |||
---|---|---|---|---|---|---|
MPEG-4 Video | 640x360 | 1.94 Mbits/sec | 664.71 MB | View | Download | |
WebM | 640x360 | 669.06 kbits/sec | 223.70 MB | View | Download | |
iPod Video | 480x270 | 522.35 kbits/sec | 174.59 MB | View | Download | |
MP3 * | 44100 Hz | 249.76 kbits/sec | 83.57 MB | Listen | Download | |
Auto | (Allows browser to choose a format it supports) |