Inference from trees: Cybersecurity
26 mins 28 secs,
48.43 MB,
MP3
44100 Hz,
249.81 kbits/sec
Share this media item:
Embed this media item:
Embed this media item:
About this item
| Description: |
Thomas
Tuesday, March 16, 2021 - 11:45 to 12:10 |
|---|
| Created: | 2021-03-17 12:51 |
|---|---|
| Collection: | Unlocking Data Streams |
| Publisher: | Isaac Newton Institute |
| Copyright: | Thomas |
| Language: | eng (English) |
| Distribution: |
World
(downloadable)
|
| Explicit content: | No |
| Aspect Ratio: | 16:9 |
| Screencast: | No |
| Bumper: | UCS Default |
| Trailer: | UCS Default |
| Abstract: | In cybersecurity we want to detect malicious activity by finding unusual patterns of behaviour. The logs of events on a system can be modelled as trees that evolve over time; the path signature provides a way to extract features from these trees by considering them as a collection of paths. We can then use these features in machine learning techniques, to distinguish different types of behaviour. |
|---|---|
Available Formats
| Format | Quality | Bitrate | Size | |||
|---|---|---|---|---|---|---|
| MPEG-4 Video | 640x360 | 223.99 kbits/sec | 43.37 MB | View | Download | |
| WebM | 640x360 | 189.27 kbits/sec | 36.67 MB | View | Download | |
| iPod Video | 480x270 | 424.84 kbits/sec | 82.25 MB | View | Download | |
| MP3 * | 44100 Hz | 249.81 kbits/sec | 48.43 MB | Listen | Download | |
| Auto | (Allows browser to choose a format it supports) | |||||