Devpost for Teams
Grow your developer ecosystem and promote your platform.
Drive innovation, collaboration, and retention within your organization.
Analyzing adverse drug reactions through data visualization and machine learning.
We use NIST-MST algorithm for differential privacy to prevent reconstruction attacks. We examine its effects on different models, and maintain a reasonable degree of accuracy while improving privacy.
At first glance, people think there are two stages to the soft pity system in Genshin Impact's wish system but there is actually only one stage of the soft pity system.
To enhance the privacy of the yellow cabs dataset, we identify critical attributes. Then, we employ Laplacian noising for continuous attributes and an ML classifier to mask categorical attributes.
How do factors such as age and sex influence how patients react to different medicines?
Our approach simultaneously preserves differential privacy and offers high levels of accuracy in synthesized data via conditional table GAN.
When the Cab company extract data, they need to preserve privacy. If the customer tip very little it could be embarrassing. Our service generate synthetic tip-to-fare ratio data to avoid this problem
Our goal is to evaluate the dangerously used drugs and the amount of doses by forming machine learning models.
Visualizing medical data to analyze the seriousness of the drug, analyzing number of cases, drug effects, visualizing using geo, mapping, bargraph improve decision-making and understand impact of it.
Investigation of the correlation between Acetaminophen and different factors, including age, sex, and co-ingestion with other drugs.
Harnessing Python, R, and Excel to safeguard taxi cab driver and passenger data.
Gaming Track - Genshin Impact
We explore 5 ways of generating data on yellow cabs that differentially-privatizes the original dataset. Methods build upon previous methods until we result in a stable dataset-generating algorithm.
To generate the synthetic dataset, we used two techniques - graphical model based estimation and generative approach and validated the tradeoff between data privacy and model performance.
Privacy is of vital importance in this big data era. Companies providing query services and ML algorithm endpoints should take care of any form of threat to their data privacy.
# Research on the framework of DP
# Implement RON-Gaussian
# Grid search for the best privacy-accuracy tradeoff
We analyzed trends in adverse effects of antihypertensive drugs and built a model to predict which adverse effect patients are most likely to experience. Our model has a 50% accurancy rate.
Patient safety is the fundamental component in healthcare services. This project aims to develop an effective model to decrease possible preventable adverse effect rate in practice.
We use a mixture model to describe how loot box rewards depend on a player's "pity score," a metric which measures the number of consecutive attempts since a player last obtained a legendary item.
In this exciting project, we will delve into the world of Genshin, a popular RPG set in a fantastical universe.
Exploring correlations between adverse effects in patients and demographic information in harder to diagnose conditions to identify biases in healthcare treatment
Using a synthetic gauss and feed forward neural network, we generate data to mimic a given dataset. Our approach follows trends and accurately extends datasets.
1 – 24 of 30