HousePrice strives to predict the rental price of an AirBnb rental unit, given some information like the city of location, neighborhood, location, room type, number of nights, number of reviews, availability etc. The approach involves using data about these parameters from insideairbnb.com, and using the transaction information available for each listing to predict the right rental price to maximize profits for the home owner.
Feedback:
Positive:
1) The utility of a such a system will be arguably very valuable for homeowners who want to start using AirBnb, and for propsective homeowners who want to invest in buying homes to rent them out as AirBnb units
2) The dataset is reliable and has extensive amount of information
3) A number of models can be used here, so the developers will learn a lot. Some linear models combined with feature engineering should be a good approach.
Negative/Questions:
1) The goal is vaguely defined - are you trying to predict the daily rate, total cost of stay, or price of buying an equivalent home to rent out as an airbnb listing. You will benefit from a more concretely defined goal
2) How do you plan on featurizing your data? How do you quantify fields like neighborhood to see if they are good or bad?
3) How do you plan on using the transaction history from calendar.csv? Why is it needed, if you already have the 'price' field in your other dataset listings.csv?
HousePrice strives to predict the rental price of an AirBnb rental unit, given some information like the city of location, neighborhood, location, room type, number of nights, number of reviews, availability etc. The approach involves using data about these parameters from insideairbnb.com, and using the transaction information available for each listing to predict the right rental price to maximize profits for the home owner.
Feedback: Positive: 1) The utility of a such a system will be arguably very valuable for homeowners who want to start using AirBnb, and for propsective homeowners who want to invest in buying homes to rent them out as AirBnb units 2) The dataset is reliable and has extensive amount of information 3) A number of models can be used here, so the developers will learn a lot. Some linear models combined with feature engineering should be a good approach.
Negative/Questions: 1) The goal is vaguely defined - are you trying to predict the daily rate, total cost of stay, or price of buying an equivalent home to rent out as an airbnb listing. You will benefit from a more concretely defined goal 2) How do you plan on featurizing your data? How do you quantify fields like neighborhood to see if they are good or bad? 3) How do you plan on using the transaction history from calendar.csv? Why is it needed, if you already have the 'price' field in your other dataset listings.csv?