Training a chain structured SVM model on the OCR dataset. This example uses the Viterbi algorithm for the decoding oracle:
spark-1.X/bin/spark-submit \
--class "ch.ethz.dalab.dissolve.examples.chain.ChainDemo" \
--master local \
--driver-memory 2G \
<examples-jar-path>
Here is the same example using more general Belief Propagation, by employing the Factorie library (Requires Factorie 1.0 Jar to be placed within dissolve-struct-examples/lib
directory):
spark-1.X/bin/spark-submit \
--class "ch.ethz.dalab.dissolve.examples.chain.ChainBPDemo" \
--master local \
--driver-memory 2G \
<examples-jar-path>