FlowRec parameters

This is a list of FlowRec-specific parameters for prequential evaluation in addition to those originally provided by Scikit-Multiflow (see here).

session_column_index (required):  The index of a data column containing session identifiers. time_column_index (optional):  The index of a data column containing timestamps. Note that they are not used by the stream to order the samples. event_column_index (optional):  The index of a data column containing indexed event types. This allows for event-specific training and evaluation. rec_triggers (optional):  List of indexed event types for which recommendations are requested (only used when event_column_index is provided). allow_repeated (default=True):  Allow/disallow previously suggested items to be recommended again in subsequent events of a session. allow_reminders (default=True):  Allow/disallow previously visited items to be recommended in subsequent events of a session, thus acting as reminders. rec_size (default=10):  The size of the recommendation list (a.k.a. cutoff). n_skip (default=0):  The number of samples to skip from the start of the stream. n_keep (default=1000):  The size of the observation window.