| clear query|facets|time |
Search criteria: position.
Results from 41 to 50 from
68 (0.315s).
|
|
|
Loading phrases to help you
refine your search...
|
|
Re: combination of features worsen the performance - Mahout - [mail # user]
|
|
... as many positive examples you have. Even then, I suspect you don't have enough data especially if you have already included data for all of your models. Your Feature A is not useful unless you...
|
|
|
Author: Ted Dunning,
2011-07-11, 22:30
|
|
|
Re: Best way to do a recommendation engine based on CLR (Click Through Rate) - Mahout - [mail # user]
|
|
...The position of users and items in the user x item occurrence matrix are interchangeable. For every solution problem that matches items to users, there is a dual solution that matches...
|
|
|
Author: Ted Dunning,
2010-07-28, 16:53
|
|
|
Re: Strange results running SGD TrainNewsGroups example - Mahout - [mail # user]
|
|
... a look again. My basic expectation is that you don't have anything going wrong. If you want to see positive only terms, you might try tweaking the ModelDissector to sort on value descending...
|
|
|
Author: Ted Dunning,
2010-12-18, 00:58
|
|
|
Re: Problem with CNB implementation? - Mahout - [mail # user]
|
|
... that the document being scored is more like the complementary class than the class being scored. This position is internally consistent at least and, since Robin A wrote that code, there is some...
|
|
|
Author: Ted Dunning,
2011-03-05, 05:44
|
|
|
Re: FeatureEncoder Question - Mahout - [mail # dev]
|
|
... divide the space into two sub-spaces, you get vastly less than half capacity in each sub-space. This behavior is related to Bloom filters where a fixed false positive rate leads to a size...
|
|
... proportional to the number of documents and the number of bits per document is proportional to the negative log of the false positive rate. The upshot is that it is better to mix everything...
|
|
|
Author: Ted Dunning,
2010-09-02, 04:24
|
|
|
Re: AdaptiveLogisticRegression.close() ArrayIndexOutOfBoundsException - Mahout - [mail # user]
|
|
... of work to find positive examples. The general technique you need is called active learning. This is where the model helps you find training data to hand tag. There are two sub-problems. One...
|
|
|
Author: Ted Dunning,
2011-05-02, 18:01
|
|
|
Re: Feedback on provided recommendations - Mahout - [mail # user]
|
|
... wind up recommending everything except what the user likes (if the negative ratings overwhelm the positive in spite of the blending parameters) or you can wind up missing some very good stuff...
|
|
|
Author: Ted Dunning,
2010-04-27, 19:53
|
|
|
Re: n-gram over-representation? - Mahout - [mail # user]
|
|
... of word A (and complementarily, any non-A word) in the first position and word B (and non-B words) in the second position. No. I was referring to my own napkin scribblings. If you expand...
|
|
|
Author: Ted Dunning,
2010-02-16, 20:18
|
|
|
Re: Order-based evalution of recommenders - Mahout - [mail # user]
|
|
... positive has a higher score (appears earlier) than a random true negative example. If you assume that all items that do not appear in your reference set are negatives, then you can iterate over...
|
|
|
Author: Ted Dunning,
2010-10-31, 03:45
|
|
|
Re: A mahout logo Revamp - Mahout - [mail # dev]
|
|
... forward. The logo will have only marginal positive impact on the project even if it is fun to decide on. On Sun, Mar 14, 2010 at 8:42 AM, Jeff Eastman wrote: ...
|
|
|
Author: Ted Dunning,
2010-03-14, 19:39
|
|
|
|
