1. Create a SequenceRecordReader instance to extract and load features from the time series data:

SequenceRecordReader trainFeaturesSequenceReader = new CSVSequenceRecordReader();
 trainFeaturesSequenceReader.initialize(new NumberedFileInputSplit(new File(trainfeatureDir).getAbsolutePath()+"/%d.csv",0,449));

2. Create a SequenceRecordReader instance to extract and load labels from the time series data:

SequenceRecordReader trainLabelsSequenceReader = new CSVSequenceRecordReader();
 trainLabelsSequenceReader.initialize(new NumberedFileInputSplit(new File(trainlabelDir).getAbsolutePath()+"/%d.csv",0,449));

3. Create sequence readers for testing and evaluation:

SequenceRecordReader testFeaturesSequenceReader = new CSVSequenceRecordReader();
 testFeaturesSequenceReader.initialize(new NumberedFileInputSplit(new File(testfeatureDir).getAbsolutePath()+"/%d.csv",0,149));
 SequenceRecordReader testLabelsSequenceReader = new CSVSequenceRecordReader();
 testLabelsSequenceReader.initialize(new NumberedFileInputSplit(new File(testlabelDir).getAbsolutePath()+"/%d.csv",0,149));|

4. Use SequenceRecordReaderDataSetIterator to feed the data into our neural network:

DataSetIterator trainIterator = new SequenceRecordReaderDataSetIterator(trainFeaturesSequenceReader,trainLabelsSequenceReader,batchSize,numOfClasses);

DataSetIterator testIterator = new SequenceRecordReaderDataSetIterator(testFeaturesSequenceReader,testLabelsSequenceReader,batchSize,numOfClasses);

5. Rewrite the train/test iterator (with AlignmentMode) to support time series of varying lengths:
   

DataSetIterator trainIterator = new SequenceRecordReaderDataSetIterator(trainFeaturesSequenceReader,trainLabelsSequenceReader,batchSize,numOfClasses,false, SequenceRecordReaderDataSetIterator.AlignmentMode.ALIGN_END);