Dublin Airbnb Market Analysis

Inspiration

We chose this because it's been a long time since we used an Airbnb (the prices are no longer attractive compared to hotels, because of rising interest rates that pushed Airbnb investors on Wall Street to urge the company to quickly become profitable), and we were curious how the company looked like 10 years ago, when it was steadily approaching its peak popularity.

What it does

Our project analyzes the 2014 Dublin Airbnb market and identifies key customer segments that can be better targeted.

How we built it

We used contacts and search data from stratascratch and clustering algorithms to identify characteristics of the customer segments.

Challenges we ran into

We tried using an OPTICS algorithm for customer segmentation, but it did not work and gave arbitrary results. So, we pivoted to using faster, but less accurate clustering algorithms (i.e., K-means and Gaussian mixture model).

Accomplishments that we're proud of

Doing our datathon and getting useful results.

What we learned

We learned hyperparameter tuning and basic data science concepts including, but not limited to, clustering and unsupervised learning.

What's next for 6044

Doing more complicated data analysis with information that is not just text (i.e., image, video, etc..)

Built With

• matplotli
• numpy
• pandas
• python
• scikit-learn
• seaborn