Luis Villazon's repositories
Mitigating-Bias-in-Heart-Disease-Prediction
We used a Random Forest Classifier on the Heart Dataset to predict heart disease. Initial results favored females, so we adjusted the decision threshold for males to 0.3, improving recall without impacting female accuracy. This ensures our AI models fairly benefit all users.
Advertising-Campaigns-AB-Testing
This research examines the effectiveness of targeted advertisements compared to public service announcements analyzing data from participants exposed to a varying number of ads. Using techniques such as Propensity Score Matching and hypothesis testing, the study aims to understand how ad exposure influences conversion rates.
Airline-Strategic-Overbooking-and-Dynamic-Pricing
The project focuses on optimizing airline revenue through strategic overbooking and dynamic pricing for coach and first-class tickets, aiming to increase profits while managing overbooking risks and customer relations.
complaints_analysis
Reclassify customer complaints based on the exact language used in the complaint