| clear query|facets|time |
Search criteria: ClusterEvaluator.
Results from 121 to 130 from
548 (1.529s).
|
|
|
Loading phrases to help you
refine your search...
|
|
Re: [mahout] labels in clustering algorythms - Mahout - [mail # user]
|
|
...Yuriy, Take a look at https://github.com/tdunning/knn to see some upcoming k-means stuff that may help you out with respect to speed. On Sat, Apr 28, 2012 at 11:19 AM, Юрий Басов...
|
|
|
Author: Ted Dunning,
2012-04-28, 19:50
|
|
|
Re: Clustering : Number of Reducers - Mahout - [mail # user]
|
|
...With potentially 5 million clusters, I would suggest a multi-pass hierarchical approach where you cluster the data and then sub-cluster the larger clusters. On Tue, Sep 20, 2011 at 12...
|
|
|
Author: Ted Dunning,
2011-09-20, 21:20
|
|
|
Re: Cluster: find medoid & its n nearest elements - Mahout - [mail # user]
|
|
...There isn't a clever way to find the medoid in Mahout. Finding the n nearest elements can be done using a Searcher. The Brute implementation should suffice. On Thu, Dec 6, ...
|
|
|
Author: Ted Dunning,
2012-12-07, 12:27
|
|
|
Re: Clustering : Number of Reducers - Mahout - [mail # user]
|
|
...On Sun, Sep 18, 2011 at 11:02 PM, Paritosh Ranjan wrote: No. It means that in most clustering algorithms the combiners do enough work that the reducers have relatively little to do. ...
|
|
|
Author: Ted Dunning,
2011-09-19, 05:17
|
|
|
Re: some new clustering code - Mahout - [mail # dev]
|
|
... of maxClusters was intended to play to the strength of this kind of 1-pass algorithm which is to find a lot of clusters that characterize the data well. The desired application is in nearest...
|
|
... neighbor search were we want to see thousands to hundreds of thousands of clusters, but I wanted to make sure that users didn't accidentally get bad results by specifying that they wanted 10...
|
[+ show more]
[- hide]
| ... clusters from a million data points. Note that a side effect of this choice is that you really need much more than 100 data points to have this algorithm make sense. The intended application... |
| ... is with datasets in the millions of rows or more. There is also a difference in that StreamingKmeans uses polymorphism to allow the clustering of points to share code with the clustering of centroids... |
|
|
Author: Ted Dunning,
2012-04-06, 20:50
|
|
|
Re: Clustering Data - Mahout - [mail # user]
|
|
...OK. This is more of a kind of time series analysis even if the horizontal axis isn't time. You need to extract features from these graphs before doing clustering. Something like extreme...
|
|
|
Author: Ted Dunning,
2011-08-16, 18:18
|
|
|
Re: Clustering Data - Mahout - [mail # user]
|
|
... can noodle around with all the clustering algorithms in the world, select one and port it into Java or find an implementation. And if you don't describe your problem in a bit more detail, we...
|
|
... can't help you. Clustering specifically and machine learning in general is domain dependent. Your graphs don't explain what your data is, what your are trying to do, what results you expect...
|
|
|
Author: Ted Dunning,
2011-08-15, 15:08
|
|
|
Re: Clustering Data - Mahout - [mail # user]
|
|
...I don't think that you need a solver for this. Just compute the smoothed derivatives at all points and then extract features from the rank statistics of the ensemble of all values. &nb...
|
|
|
Author: Ted Dunning,
2011-08-17, 14:29
|
|
|
Re: Evaluation of different recommendation algorithms for 12.000 user data set - Mahout - [mail # user]
|
|
...Filtering recommendations lists is incredibly important. What you are doing is pretty straightforward with post-processing of the recommended list. Other things that I often reco...
|
|
|
Author: Ted Dunning,
2011-11-21, 18:08
|
|
|
Re: Evaluation of different recommendation algorithms for 12.000 user data set - Mahout - [mail # user]
|
|
...Your product is subject to seasonality constraints (which teas are likely right now) and repeat buying. I would separate out the recommendation of repeat buys from the separation of ne...
|
|
|
Author: Ted Dunning,
2011-11-21, 15:20
|
|
|
|
