Hey everyone for my thesis I am conducting research regarding ML practises within the revenue management department of hotels and how it affects the room rate.
The goal of my paper is the following: To understand how both the insignificant and the significant factors influence the outcome of the ML model with regards to the room rate of a hotel.
I have identified the insignificant and significant factors (See attached photo) that influence the ML outcome with regards to the room rate of a hotel, however I am struggling with how to actually achieve my research goal.
My connected hypotheses are:
(H0): The insignificant factors do not influence the outcome of the ML model with regards to the room rate of a hotel compared to the significant factors.
(H1): The insignificant factors do influence the outcome of the ML model with regards to the room rate of a hotel compared to the significant factors.
Have you already tried using a ML model (Decision Tree, Random Forest and fit it with your variables to see the performance based on the chosen features?
br
I don’t know if I understood well your problem, but you are looking to validate that the insignificant factors that you choose won’t influence the outcome of your ML whereas the significant will.
May I ask how did you identified the insignificant and significant variables? Building a model usually provides you with information about (in)significant variables so not sure what your exact goal is. Also I’m a bit confused with your hypothesis. By definition the insignificant factors do not influence the outcome so there’s nothing to test really here. However if you want to check that variables you think are insignificant are indeed insignificant then it makes sense building a model. Also it doesn’t seem so important to accept or reject null hypothesis in this case but rather finding out which variables are (in)significant. Or actually making one hypothesis for every variable you think is insignificant. Hope this helps!