Differences
This shows you the differences between two versions of the page.
| Both sides previous revision
Previous revision
Next revision
|
Previous revision
|
courses:mapreduce-tutorial:step-15 [2012/01/26 00:06]
straka |
courses:mapreduce-tutorial:step-15 [2012/01/29 16:40] (current)
straka |
| Implement the [[http://en.wikipedia.org/wiki/K-means_clustering#Standard_algorithm|K-means clustering algorithm]]. You can use the following data: | Implement the [[http://en.wikipedia.org/wiki/K-means_clustering#Standard_algorithm|K-means clustering algorithm]]. You can use the following data: |
| ^ Path ^ Number of points ^ Number of dimensions ^ Number of clusters ^ | ^ Path ^ Number of points ^ Number of dimensions ^ Number of clusters ^ |
| | ''/home/straka/hadoop/example-inputs/points-small'' | 10000 | 50 | 50 | | | ''/net/projects/hadoop/examples/inputs/points-small'' | 10000 | 50 | 50 | |
| | ''/home/straka/hadoop/example-inputs/points-medium'' | 100000 | 100 | 100 | | | ''/net/projects/hadoop/examples/inputs/points-medium'' | 100000 | 100 | 100 | |
| | ''/home/straka/hadoop/example-inputs/points-large'' | 500000 | 200 | 200 | | | ''/net/projects/hadoop/examples/inputs/points-large'' | 500000 | 200 | 200 | |
| | |
| | When dealing with iterative algorithms, each iteration is usually implemented as one Hadoop job. The Hadoop ''input_path'' should contain the input data and each mapper should also read the current clusters. The reducers are used to aggregate the data and output new cluster centers. A controlling script should take care of executing Hadoop jobs and stopping the iteration when the algorithm converges. |
| | |
| | ---- |
| | |
| | <html> |
| | <table style="width:100%"> |
| | <tr> |
| | <td style="text-align:left; width: 33%; "></html>[[step-14|Step 14]]: N-gram language model.<html></td> |
| | <td style="text-align:center; width: 33%; "></html>[[.|Overview]]<html></td> |
| | <td style="text-align:right; width: 33%; "></html><html></td> |
| | </tr> |
| | </table> |
| | </html> |
