NYC Rental Listing Popularity Study

Posted on

NYC Rental Listing Popularity Study (Fall 2018)

This project focus on designing a more reasonable listing mechanism that can predict the interest level (high, medium, low) of rental apartments based on their numerical and descriptive information. We approached it using logistic regression, random forest, and kNN. The results show that random forest has the best performance in predicting the popularity of rental apartments.

The project is the final project for NYU DS-GA 1007 Programming for Data Science

Data

Data comes from kaggle Two Sigma Connect: Rental Listing Inquiries.
Also uploaded, simply download everything and run the Jupyter Notebooks.

Code

Source Code
Packeges used: pandas, numpy, scikit-learn, etc.

All the code is in the 07 project Jupyter Notebook, sections are Data Importing, Data Cleaning, Model Selection and Implement, and some Data VIsualization in the end.

Authors