Project information
- Category: Machine Leraning Classification and Clustering for bank accounts
- Client: Universidad de Buenos Aires
- Project date: september-december 2023
- Project URL: Link to project
Description
For this project, I had to participate in three Kaggle competences, which consisted in trying to predict
which bank acccount was about to leave the bank in two months.
In the first competence (kaggle, github),
we were limited by using only CART. In the second (kaggle, github), we weere allowed to use Gradient Boosting of Regression Trees.
In the final one (kaggle, github), we were also allowed to use ensamble models. For the two last competences, we used Google Cloud.
Additionally, I had to design two pitches in video format. In the first one, I had to present a segmentation of clients using
Random Forest for hierarchical clustering. In the second one, I presented the results of the ensamble model for the third competence.