saving the DFA version of model when the postprocessing is requested

I set up the input argument of PRINS as follows:
```shell script
python run_PRINS.py Hadoop dataset_example/Hadoop/Hadoop_preprocessed_logs.csv --det heuristic
```
According to the instructions of the file, it should post-process the model with the heuristic technique and save the model.

However, the output model in PDF and dot format is in the NFA version and not post-processed.