Career Accelerator Program
Used data from a real career accelerator program (name not disclosed due to confidentiality) to determine whether a fellow will get a job and to predict the length of the job search using supervised machine learning. F1 score of 96.6 percent achieved.
Charity ML Donors Project
Helped CharityML, a non-profit, classify whether individuals earn more than or less than 50K so that they can target their campaign strategies for donation using supervised machine learning. Accuracy of 81 percent achieved.
Image Classification Project
Trained a deep neural network to do image classification on flowers using PyTorch and performed transfer learning using AlexNet. Built a Python module to automate training and prediction based on user input.
Identifying Customer Segments
Used data from Bertelsmann Arvato Analytics to identify customer segments using k-means clustering and Principal Component Analysis techniques to direct marketing campaigns to increase the expected revenue of the firm.