PROJECTS
While studying at General Assembly I was able to design several projects that focused on different aspects of the data science workflow. Click through for more information and to read my code.
Latest Projects
Neural Network Plant Classifier
Is it possible to help new South African farmers increase their yield by proactively pruning away the weeds that grow kill crops?" Using a neural network, I was able to classify photos with 85% accuracy. A seven-layer Convolutional Neural Netwas the backbone of an application that can take in a full directory of images and make predictions quickly and efficiently while telling the user the location of weeds in the crop field.
Develop a formula, using existing pricing databases, to calculate property damage caused by natural phenomena within a specific zip code or area.
Is it possible to classify a story as fact or fiction with machine learning? My research says yes. Using an API and a web scraping function I brought in almost 5000 posts total from two different sections of the forum Reddit.com, and classified them as fact or fiction.
Testimonials
“The current process for finding government contracts was complicated and inefficient. Greg created a custom match rate and contract rating which was key in providing more refined searches. ”
Arielle Miro, Government Hackathon Team member