Add as an input a single column from the specified RecordReader / SequenceRecordReader
The assumption is that the specified column contains integer values in range 0..numClasses-1;
this integer will be converted to a one-hot representation
An an output, where the output is taken from a single column from the specified RecordReader / SequenceRecordReader
The assumption is that the specified column contains integer values in range 0..numClasses-1;
this integer will be converted to a one-hot representation (usually for classification)
Constructor for classification, where:
(a) the label index is assumed to be the very last Writable/column, and
(b) the number of classes is inferred from RecordReader.getLabels()
Note that if RecordReader.getLabels() returns null, no output labels will be produced
RecordReaderMultiDataSetIterator: A MultiDataSetIterator for data from one or more RecordReaders and SequenceRecordReaders
The idea: generate multiple inputs and multiple outputs from one or more Sequence/RecordReaders.
When dealing with time series data of different lengths, how should we align the input/labels time series?
For equal length: use EQUAL_LENGTH
For sequence classification: use ALIGN_END
Sequence record reader data set iterator.
Given a record reader (and optionally another record reader for the labels) generate time series (sequence) data sets.
Supports padding for one-to-many and many-to-one type data loading (i.e., with different number of inputs vs.
For use with timeseries trained with tbptt
In a given minbatch, shorter time series are padded and appropriately masked to be the same length as the longest time series.