I'm looking for details about techniques for evaluation of classifiers.
Hi.
the following paper provides a very comprehensive survey of typical performance/evaluation measures typical in classification:
http://rali.iro.umontreal.ca/rali/sites/default/files/publis/SokolovaLapalme-JIPM09.pdf
It comprises for instance, Average Accuracy, Error Rate, Precision, Recall, F-score,
Precision, Recall. And many more.
Hi.
the following paper provides a very comprehensive survey of typical performance/evaluation measures typical in classification:
http://rali.iro.umontreal.ca/rali/sites/default/files/publis/SokolovaLapalme-JIPM09.pdf
It comprises for instance, Average Accuracy, Error Rate, Precision, Recall, F-score,
Precision, Recall. And many more.
In fact, I'm looking for techniques for selection data sets of trainning and tests in differents kinds of database to build classifiers; i.e; what strategy is more adequate for a type of datasets like a dataset with few instances or a dataset with many instances and/or multiclasses.(k-fold-cross-validation, LOO,...)
See this paper:
Gama, Joao, and Pavel Brazdil. "Characterization of classification algorithms." Progress in Artificial Intelligence. Springer Berlin Heidelberg, 1995. 189-200.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.47.5612&rep=rep1&type=pdf
I think your question is cae sensitive....either physical rules and data mining and cross validation techniques should be cnsidered o selecting the input configurations. I have done such works for hydrological issues (see my papers). Nevertheless, if you look for cross validation and data management scenarios papers, I can provide!
Jalal Shiri
Jalal Shiri
That's correct. I'll see yours papers and tell you about.
best regards
- Badges
- Science topic
- Similar topics
- Data Mining
- Classification