Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision | |||
|
courses:rg:2012:distributed-perceptron [2012/12/16 23:40] machacek |
courses:rg:2012:distributed-perceptron [2012/12/16 23:44] (current) machacek |
||
|---|---|---|---|
| Line 22: | Line 22: | ||
| === 4.1 Parameter Mixing === | === 4.1 Parameter Mixing === | ||
| + | |||
| + | * First attempt to map-reduce algorithm | ||
| + | * Divide the training data into shards | ||
| + | * In MAP step, train a perceptron on each shard | ||
| + | * In REDUCE step, average the trained weight vectors | ||
| === 4.2 Iterative Parameter Mixing === | === 4.2 Iterative Parameter Mixing === | ||