en-de
en-es
en-fr
en-pt
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
How to Predict Sequences with Bayes, MDL and Experts
Published on Feb 25, 20073701 Views
Related categories
Chapter list
Minimum Description Length: Contents00:01
Minimum Description Length: Abstract01:55
How to Choose the Prior?01:57
Indi®erence or Symmetry Principle04:36
The Maximum Entropy Principle ...07:21
Occam's Razor | The Simplicity Principle07:23
Pre¯x Sets/Codes07:27
Proof of the Kraft-Inequality09:37
Priors from Pre¯x Codes12:10
A Universal Choice of » and M15:18
The Minimum Description Length Principle20:11
Predict with Best Model23:54
Application: Sequence Prediction29:04
Application: Regression / Polynomial Fitting31:40
MDL Solution to Polynomial Fitting33:50
MDL Polynomial Fitting: Determine Degree d37:27
Minimum Description Length: Summary42:00
The Similarity Metric: Contents44:06
The Similarity Metric: Abstract44:46
Kolmogorov Complexity44:49
The Universal Similarity Metric47:45
Tree-Based Clustering52:32
Genomics & Phylogeny: Mammals53:24
Genomics & Phylogeny: Mammals58:20
Genomics & Phylogeny: SARS Virus and Others01:01:36
Genomics & Phylogeny: SARS Virus and Others01:01:50
Classi¯cation of Di®erent File Types01:02:37
Classi¯cation of Di®erent File Types01:03:09
Language Tree (Re)construction01:04:01
sheme01:04:29
Classify Music w.r.t. Composer01:08:38
Classify Music w.r.t. Composer01:09:27
Further Applications01:10:19
The Clustering Method: Summary01:10:41
Prediction with Expert Advice: Contents01:12:16
Prediction with Expert Advice: Abstract01:26:43
Prediction with Expert Advice (PEA) - Informal01:26:45
Prediction with Expert Advice (PEA) - Setup01:27:34
Goals01:27:49
Best Expert in Hindsight (BEH)01:29:34
Weighted Majority (WM)01:29:46
Follow the Perturbed Leader (FPL)01:30:14
Regret Bounds for n < 1 and ki = ln n01:30:54
Regret Bounds for n = 1 and general ki01:32:12
Some more FPL Results01:32:21
PEA versus Bayes Bounds { Formal01:32:23
Naive Ansatz: Follow the Leader (FL)01:33:23