| GreenComputerVision |
Deep Learning, Computer Vision, Azure Stack, Python |
Looking into ACC, AUC, and cost variations between 6 PyTorch models using the AzureML stack and a NVIDIA T4 GPU. Using DenseNet121, DenseNet169, DenseNet201, ResNet18, ResNet50, and InceptionV3 pretrained on ImageNet and fine-tuned to the MedMNIST pediatric pneumonia dataset, performance measurements and compared. |
| Pm25_HumanHealth |
Data Wrangling, Regression, Statistical Analysis, Data Visualization, R |
Combining two datasets from University of Washington Institute for Health Metrics and Evaluation and US Center for Disease Control which both use county level observations, to inspect any relationship between airborne PM2.5 pollution concentration and human mortality. |
| Fetal_Health |
Machine Learning, Multiclass Classification, Model Selection, Feature Extraction, Python |
Find key features for medical professionals to monitor in a clinical setting to assess fetal health via cardiotocograms. |
| Falling_Sky |
API Interaction, Python |
Query the NASA API to gather information available about Near Earth Objects (NEO) and then calculate details about a potential impact with Earth. Organize the details and supplementary information into a data package for a UI display. |
| RecreationFinder |
API Interaction, Applied Mathematics, PoC/Prototyping, Python |
Create a Proof of Concept for an app which takes in a user supplied address (ex: "257 E Center Street Moab UT") and finds the nearest national park. This should not just be the name of a park but also contain information about the park selected to provide a better user experience. |
| CheatDetection |
API Interaction, Data Curation, Statistical Analysis, R, Python |
Generate a player performance dataset and statistically identify accounts which are likely cheating during PUBG tournament play. |