Midterm Peer Evaluation (xp33)

[XinranPan]

Under Data Cleaning, which features did you actually analyze which have/may have a certain impact on price? For your final report, it would be helpful to add information about your clean up in this section of the paper; briefly state the specific features which have a certain impact on price, the highly correlated and uncorrelated features. It is great that you added more details describing how the features were narrowed down in the following sections, so in that case, maybe include something along the lines of “In order to better understand the model, we first cleaned up our data, as specified in the following sections.” This way, the reader won’t get misled and after reading the Data Cleaning section, think that your group didn’t care to elaborate when in fact, you spend a huge chunk of the paper explaining the process used to narrow down your features. Also, briefly describe where you obtained the data set from.

Use smaller bins for your histogram. This could help show clearer distribution and spread of the data.

Impressive use of both LASSO and ridge regression.

Will you be further exploring your pricing predictions through analysis on both model 1 and 2? Seem from the Future Work section you wish to use Model 1, although I am not certain from reading this.

In general, super interesting and very useful to any Airbnb host. You’ve made good progress and have even done a few sophisticated regressions, like LASSO and ridge. Most of the suggestions I have for you are related to the presentation of the project. Be sure to state everything simply and clearly so readers who have never heard of your project will have an easy time understanding from beginning to end. I wonder if you could actually get a few hosts to test your finished project, run a controlled experiment to test your profit predictions, and compare those results with how much revenue your model states.