Feature Bundles

[abstractqqq]

Let's brainstorm some useful feature bundles (features that are likely to be generated together).

Some common ideas are:

1. Diagnostic features. Features that detect anomaly in time seres.
2. Pattern detection features.
3. Features specific in certain domains.
4. Features that are commonly used together in a variety of domain.
5. etc...

@MathieuCayssol @topher-lo @vienneraphael

[abstractqqq]

For common diagnostic, e.g. stability of time series, the features below can be grouped.

• absolute_maximum
• count_above_mean
• count_below_mean
• longest_strike_above_mean
• longest_strike_below_mean
• mean_change
• mean_abs_change
• mean_n_absolute_max
• number_crossings
• range_count
• ratio_beyond_r_sigma
• coefficient of variation

I am thinking about adding a few features not in Tsfresh

• greatest reversion from peak/max, value and percentage
• max/min_change
• longest downward/upward strike
• average/median downward/upward strike

[topher-lo]

Your suggested features are great for finance

[baggiponte]

Compatibility with ML tracking tools like mlflow?