0
$\begingroup$

I have the following time-series data:

date        volatility contractType moneyness
2018-01-02  0.656351   A            89.860333
2018-01-02  0.561393   A            91.383390
2018-01-02  0.463378   A            92.958966
2018-01-02  0.361770   A            94.589825
2018-01-02  0.319940   A            95.258304
...         ...        ...          ...
2022-03-10  0.900534   D            120.655385
2022-03-10  0.939075   D            121.960000
2022-03-10  0.976861   D            127.229091
2022-03-10  1.022399   D            138.952000
2022-03-10  1.875746   D            139.760000
  • Type of contract is categorical (A,B,C,D)
  • Moneyness ranges from 50-150, and is continuous. To reduce dimensionality, I could also bin it into ~10 categorical bins.

Which ML model should I use to obtain a trained function $\hat{f}(date, contractType, moneyness) = volatility $ ?

I am confused about the type of model and data I should use to obtain a trained function of time, and exogenous variables $ f(t, x_1, x_2, .. x_n) = y$.

Should $x's$ be categorical ? Can I leave moneyness as continuous?

$\endgroup$

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.