Fingertip pulse oximeters are the current standard for estimating blood oxygen saturation without a blood draw, both at home and in healthcare settings. However, pulse oximeters overestimate oxygen saturation, often resulting in ‘hidden hypoxemia’: a patient has hypoxemia (dangerously low oxygen saturation), but the oximeter returns a healthy oxygen value. Unfortunately, oximeter overestimation of oxygen saturation is exacerbated for patients with darker skin tones due to light-based oximeter technology. This results in Black patients experiencing hidden hypoxemia at twice the rate of white patients. By combining pulse oximeter readings (SpO2) with additional patient data, we develop improved methods for estimating arterial blood oxygen saturation (SaO2) and identifying Hidden Hypoxemia. The predictions of our models are more accurate than pulse-oximeter readings alone, and remove the systematic racial inequity inherent in the current medical practice of using oximeter readings alone.

Predict the biological function of proteins based upon their amino acid sequence and other publicly available data.

This project would double as a team submission to the Fifth Critical Assessment of Functional Annotation (CAFA 5) competition - similar to the Critical Assessment of Structure Prediction (CASP 14) competition in which AlphaFold2 gained prominence (in 2021).

Here is the Kaggle competition (submissions are open for the next 3 months): https://www.kaggle.com/competitions/cafa-5-protein-function-prediction/
(1st prize is $15,000... lol)

How to make money in LA with Airbnb? Being a host has never been easy since they need to provide a quality experience for guests while considering the profit margin. We aim to alleviate their burdens by streamlining the decision-making process. We explore and analyze tens of thousands of listings in LA for insights. Then, we combined structural modeling and machine learning to customize pricing for new listings based on property locations, features, ownership, and substitution between nearby listings to maximize the host’s profit. The structural model helps capture supply & demand dynamics and machine learning helps capture the consumer consideration set (hotspot market). We also use machine learning for price prediction to utilize rich features better. Furthermore, we recommend areas with the best rate of return for potential hosts and suggest possible amenities to become popular in Airbnb.