What features/model to use for data with Hysteresis?

I have made a pressure sensor that when graphing Conductance vs Pressure (disregard the actual values in graph), has the following behaviors:

1. First pressed it has one trendline
2. Afterwards when decreasing pressure it shows hysteresis
3. New trendline when increasing pressure.
4. All new cycles of increase/decrease pressure follow (more or less) the lines of 2 and 3.

Usually we fit the data with a linear equation. But was wondering if it made sense to use ML to fit the data, feeding in the derivative as an additional feature (thinking of linear regression).

The end goal is to use Conductivity (feature) to predict Pressure (label) , reducing the error by accounting for the hysteresis.

What features should I Use? What model is best?