PAMAP2 dataset preprocessed v0.2.0

# Processed PAMAP2 dataset
This dataset is based on the [PAMAP2 Dataset for Physical Activity Monitoring](https://archive.ics.uci.edu/ml/datasets/PAMAP2+Physical+Activity+Monitoring).

The data is processed with the code from [this script](https://github.com/NLeSC/mcfly-tutorial/blob/master/tutorial/tutorial_pamap2.py), with the following function call:

```python
columns_to_use = ['hand_acc_16g_x', 'hand_acc_16g_y', 'hand_acc_16g_z',
                'ankle_acc_16g_x', 'ankle_acc_16g_y', 'ankle_acc_16g_z',
                  'chest_acc_16g_x', 'chest_acc_16g_y', 'chest_acc_16g_z']
exclude_activities = [9, 10, 11, 18, 19, 20, 0]
outputpath = tutorial_pamap2.fetch_and_preprocess(directory_to_extract_to,
                                        columns_to_use,
                                        exclude_activities=exclude_activities,
                                        val_test_size=(100, 1000))
```

## References
A. Reiss and D. Stricker. Introducing a New Benchmarked Dataset for Activity Monitoring. The 16th IEEE International Symposium on Wearable Computers (ISWC), 2012.