OpenANN
1.1.0
An open source library for artificial neural networks.
|
Data set interface. More...
#include <DataSet.h>
Public Member Functions | |
virtual | ~DataSet () |
virtual int | samples ()=0 |
Number of instances. More... | |
virtual int | inputs ()=0 |
Input dimensions of instances. More... | |
virtual int | outputs ()=0 |
Output dimensions of instances. More... | |
virtual Eigen::VectorXd & | getInstance (int n)=0 |
Get the input of the ith instance. More... | |
virtual Eigen::VectorXd & | getTarget (int n)=0 |
Get the output of the ith instance. More... | |
virtual void | finishIteration (Learner &learner)=0 |
This function is called after an iteration of the optimization algorithm. More... | |
Data set interface.
A data set can be either a training, validation or test set. Supervised data sets provide instances and targets. Unsupervised data sets might provide only instances.
See Creating a data set.
|
inlinevirtual |
|
pure virtual |
This function is called after an iteration of the optimization algorithm.
It could log results, modify or extend the data set or whatever.
learner | learned model |
Implemented in BCIDataSet, OpenANN::DataSetView, OpenANN::DirectStorageDataSet, OpenANN::WeightedDataSet, and TwoSpiralsDataSet.
|
pure virtual |
Get the input of the ith instance.
n | number of instance |
Implemented in BCIDataSet, OpenANN::DataSetView, OpenANN::DirectStorageDataSet, OpenANN::WeightedDataSet, and TwoSpiralsDataSet.
|
pure virtual |
Get the output of the ith instance.
n | number of instance |
Implemented in BCIDataSet, OpenANN::DataSetView, OpenANN::DirectStorageDataSet, OpenANN::WeightedDataSet, and TwoSpiralsDataSet.
|
pure virtual |
Input dimensions of instances.
Implemented in BCIDataSet, OpenANN::DataSetView, OpenANN::DirectStorageDataSet, OpenANN::WeightedDataSet, and TwoSpiralsDataSet.
|
pure virtual |
Output dimensions of instances.
Implemented in BCIDataSet, OpenANN::DataSetView, OpenANN::DirectStorageDataSet, OpenANN::WeightedDataSet, and TwoSpiralsDataSet.
|
pure virtual |
Number of instances.
Assumes that the data set has a fixed size, at least for one iteration of the optimization algorithm.
Implemented in BCIDataSet, OpenANN::DataSetView, OpenANN::DirectStorageDataSet, OpenANN::WeightedDataSet, and TwoSpiralsDataSet.