/
leland-mcinnes-topological-techniques-for-unsupervised-learning-pydata-la-2019.json
18 lines (18 loc) · 1.77 KB
/
leland-mcinnes-topological-techniques-for-unsupervised-learning-pydata-la-2019.json
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
{
"description": "Many topics in unsupervised learning can be viewed as dealing with the\nrelative geometry of data. In mathematics, topology and homotopy theory\nare the fields that deal with similar kinds of questions. Using ideas,\ntechniques, and language from topology can prove fruitful for\nunsupervised learning. This talk will introduce you to the ideas and\nintuitions for this, and provide meaningful examples.\n\nMany topics in unsupervised learning can be viewed as dealing with the\nrelative geometry of data. In mathematics, topology and homotopy theory\nare the fields that deal with similar kinds of questions. Using ideas,\ntechniques, and language from topology can prove fruitful for\nunsupervised learning. This talk will look at how topological approaches\ncan be brought to bear upon unsupervised learning problems as diverse as\ndimension reduction, clustering, anomaly detection, word embedding, and\nmetric learning. Through the lens and language of topology and category\ntheory we can draw common threads through all these topics, pointing the\nway toward new approaches to these problems. By focusing on broad ideas\nand intuitions, and working with example uses, you don't need a\nbackground in topology to understand the approach. I hope to convince\nyou that topological approaches offer a rich and growing field of\nresearch for unsupervised learning.\n",
"duration": 2426,
"language": "eng",
"published_at": "2019-12-23T21:03:45.000Z",
"recorded": "2019-12-04",
"speakers": [
"Leland McInnes"
],
"thumbnail_url": "https://i.ytimg.com/vi/7pAVPjwBppo/hqdefault.jpg",
"title": "Learning Topology: Topological Techniques for Unsupervised Learning",
"videos": [
{
"type": "youtube",
"url": "https://www.youtube.com/watch?v=7pAVPjwBppo"
}
]
}