Predicting Airbnb Satifaction ratings in Seattle region using Decision Forest
To understand Seattle Airbnb customer preferences better by building a model that could predict satisfactions ratings.
- Identified tourist friendly listings using the location of each listing relative to a list of tourist attractions
- Built models deploying Decision Forest, Boosted Decision Tree and Logistic Regression
- Compared accuracy measures using P/R and ROC curves to determine the most accurate model to predict satisfaction ratings
- Further analyzed results to understand model feature importance int he decisions of Airbnb users in the Seattle region
