template, typename WT = arma::rowvec>
class mlpack::cv::MetaInfoExtractor< MLAlgorithm, MT, PT, WT >
MetaInfoExtractor is a tool for extracting meta information about a given machine learning algorithm.
It can be used to automatically extract the type of predictions and weights (if weighted learning is supported), whether the machine learning algorithm takes a DatasetInfo parameter or a numClasses parameter.
The following assumptions are made about the machine learning algorithm.
- All needed information can be extracted from signatures of Train methods.
- The machine learning algorithm contains either only non-templated Train methods or only templated ones.
- Train methods that can be used for extraction of needed information should be distinguishable by a number of arguments (for more information read discussion in https://github.com/mlpack/mlpack/issues/929).
- If weighted learning is supported, passing weights is an option rather than a requirement.
- Template Parameters
|MLAlgorithm||A machine learning algorithm to investigate. |
|MT||The type of data. |
|PT||The type of predictions (should be passed when the predictions type is a template parameter in Train methods of MLAlgorithm). |
|WT||The type of weights (should be passed when weighted learning is supported, and the weights type is a template parameter in Train methods of MLAlgorithm). |
Definition at line 272 of file meta_info_extractor.hpp.