|
|
|
|
|
|
 |
 |
|
Review of Prediction and Data Analysis Algorithms
|
|
|
|
|
|
| Organization: | Sezmi Corporation |
|
|
|
|
|
|
Wolfram Technology Conference 2013
|
|
|
|
|
|
Champaign, Illinois, USA
|
|
|
|
|
|
In this talk we are going to review machine learning and prediction algorithms using Mathematica implementations. The algorithms considered are: decision trees and forests, naive Bayesian classifiers, associative rules learning, bi-sectional clustering, non-negative matrix factorization, prefix trees, static and dynamic ranking of linked items using Markov chains, a sparse linear algebra recommender for histories over item*metadata relations.
We are going to look into couple of classifications of these algorithms and into utilization patterns for their application. All algorithms are demonstrated with relevant examples. Future directions and extensions would be discussed.
The implementations of the discussed algorithms can be found at the GitHub project MathematicaForPrediction.
|
|
|
|
|
|
|
|
|
|
|
|
http://www.wolfram.com/events/technology-conference/2013
|
|
|
 |
|
|
| ReviewOfPredictionAndDataAnalysisAlgorithms.nb (2.7 MB) - Mathematica Notebook |
|
|
|
|
|
|
|
| | | | | |
|
For the newest resources, visit Wolfram Repositories and Archives »
