leon.bottou.org

User Tools

Site Tools

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
talks [2007/08/17 12:03]
leonb 		talks [2022/04/19 11:00] (current)
leonb
Line 1: Line 1:
 ====== Talks ====== ====== Talks ======
  
-This page links the slides of my most significant lectures. +This page contains pointers to my most significant lectures.
-All the slides are encoded in [[:research:djvu|DjVu]] format. +
-You might have to install the [[http://www.djvuzone.org/download|DjVu browser plugin]]  +
-to see them.+
  
-=== ===+All the slides are available under both the PDF and DjVu formats.
  
-===== The BackPropagation CookBook =====  +  * [[talks/cookbook|The BackPropagation CookBook (NIPS Workshop 1996)]] 
- +  * [[talks/gtn|Graph Transformer Networks (ICML Workshop 2001)]] 
-{{talk-tricks.png?130 }} +  * [[talks/djvu|DjVu: Scanned Documents on the Web (2000)]] 
-This lecture is co-authored with [[http://yann.lecun.com|Yann Le Cun]]. +  * [[talks/mlss|Online Learning and Stochastic Approximations (MLSS Tuebingen 2003)]] 
-It took place at the 1996 NIPS Workshop  +  * [[talks/largescale|The Tradeoffs of Large Scale Learning (NIPS Tutorial 2007)]] 
-[[http://www.willamette.edu/~gorr/nipsws.htm|Tricks of the Trade]] +  * [[talks/onepass|Stochastic Algorithms for One Pass Learning (NIPS Workshop 2011)]] 
-organized by [[http://ida.first.fraunhofer.de/~klaus|Klaus-Robert Müller]] and [[http://www.willamette.edu/~gorr|Genevieve Orr]]. +  * [[talks/counterfactuals|Counterfactual Reasoning and Learning Systems (2012)]] 
- +  * [[talks/mlss13|Multilayer Networks [no longer old fashioned!] (2013)]] 
-  * See [[http://leon.bottou.org/slides/tricks/index.djvu|the slides (djvu 199KB)]] [[http://leon.bottou.org/slides/tricks/tricks.pdf|(pdf 2.7MB)]]. +  * [[talks/perceptrons|Perceptrons Revisited (AAAI Spring Symposium 2015)]] 
-  * See the corresponding [[:papers:lecun-98x|book chapter]]. +  * [[talks/2challenges|Two big Challenges in Machine Learning (ICML 2015)]] 
- +  * [[talks/invariances|Learning Representation with Causal Invariance (ICLR 2019)]]
-=== === +
- +
-===== Graph Transformer Networks =====  +
- +
-{{talk-gtn.png?130 }} +
-This lecture describe Graph Transformer Networks  +
-It took place at the 2001 ICML workshop [[http://web.engr.oregonstate.edu/~tgd/ml2001-workshop|Machine Learning for Spatial and Temporal Data]] organized by [[http://web.engr.oregonstate.edu/~tgd|Tom Dietterich]]. +
-[[:papers:bottou-97|Graph Transformer Networks]] are one of the most powerful and successful method for learning sequential data. About 10% to 20% of the checks written in the U.S. since 1996 have been processed by a Graph Transformer Network. Graph Transformer Networks are related to [[http://www.inference.phy.cam.ac.uk/hmw26/crf|Conditional Random Fields]] but have variable geometry and non-linear energies. +
- +
-   * See [[http://leon.bottou.org/slides/gtn/index.djvu|the slides (djvu 234KB)]] [[http://leon.bottou.org/slides/gtn/gtn.pdf|(pdf 2.4MB)]]. +
-   * See [[:papers:bottou-97|a short paper]] or [[:papers:lecun-98h|a long paper]]. +
-    +
-=== === +
- +
-===== DjVu: Scanned Documents on the Web ===== +
- +
-{{talk-djvu.png?120 }} +
-[[http://www.djvuzone.org|DjVu]] is a document compression system that allows the distribution of scanned documents on the web.  DjVu files are //very compact//. A typical 300dpi bitonal page takes 10-15KB. A typical 300dpi color page takes 40-60KB.  The presentation discusses the main technical innovations that made DjVu possible. +
- +
-  * See [[http://leon.bottou.org/slides/djvu/index.djvu|the slides (djvu 417KB)]] [[http://leon.bottou.org/slides/djvu/djvuslides.pdf|(pdf 4.9MB low quality)]]. +
-  * See the [[:research:djvu|DjVu research page]] and the [[:projects:djvulibre|DjVu software page]]. +
- +
-=== === +
- +
-===== Online Learning and Stochastic Approximations =====  +
- +
-{{talk-mlss.png?200 }} +
-This four part lecture was given at the  +
-[[http://www.mlss.cc|Machine Learning Summer School]]  +
-held in [[http://www.kyb.tuebingen.mpg.de/mlss04/mlss03|Tübingen in 2003]] +
-organized by [[http://ml.typepad.com/about.html|Olivier Bousquet]], +
-[[http://www.kyb.mpg.de/~bs|Bernhard Schölkopf]] and +
-[[http://www.ipsi.fraunhofer.de/mine/en/people/luxburg|Ulrike von Luxburg]]. +
-The lecture discusses Stochastic Approximations and in particular [[wp>Stochastic_gradient_descent|Stochastic Gradient Descent]] applied to online learning algorithms. +
- +
-  * See [[http://leon.bottou.org/slides/mlss/part1.djvu|part 1: Framework (djvu 157KB)]] [[http://leon.bottou.org/slides/mlss/part1.pdf|(pdf 2.1MB)]]. +
-  * See [[http://leon.bottou.org/slides/mlss/part2.djvu|part 2: Cookbook (djvu 459KB)]] [[http://leon.bottou.org/slides/mlss/part2.pdf|(pdf 2.7MB)]]. +
-  * See [[http://leon.bottou.org/slides/mlss/part3.djvu|part 3: Convergence and Martingales (djvu 38KB)]] [[http://leon.bottou.org/slides/mlss/part3.pdf|(pdf 1.3MB)]]. +
-  * See [[http://leon.bottou.org/slides/mlss/part4.djvu|part 4: Optimal online algorithms (djvu 55KB)]] [[http://leon.bottou.org/slides/mlss/part4.pdf|(pdf 1.4MB)]]. +
- +
-  * See the corresponding [[:papers:bottou-mlss-2004|book chapter]]. +
- +
-=== ===+
  
talks.1187366630.txt.gz · Last modified: 2007/08/17 12:03 by leonb

Page Tools