Dynamic Pricing Resorts
I’m dealing with a problem in a venture of knowledge from a number of accommodations (primarily time collection and time collection worth knowledge). The problem is to construct a dynamic pricing mannequin for every lodge, that may maintain observe of market demand in addition to particular person lodge specifics. I’m considering of utilizing a Bayesian mannequin or utilizing ensemble studying, with one LSTM for every lodge and 1 “general mannequin”. I may actually use some opinions, steering, ideias, recommendation, no matter you need to provide, in attempting to know what’s the “greatest“ mannequin and methodology on this scenario. What do you guys suppose? Any assist is actually useful
Comment ( 1 )
Do you have enough data to use LSTM models on individual accomodations? If you do that’s great, I’m just wondering as I’m in the travel industry myself (not hotels) and the credible interval on individual item bookings is so big as to be useless for prediction.
I would probably personally go with something like XGB and use booked % of epsilon nearest neighbours as a measure of individualised demand.