Projects created during the Data Scientist Nanodegree on Udacity
Evaluated and optimized several different supervised learners to optimize the donation yield of a fictious charity. The donors with the highest potential donation yield were selected in order to reduce the total numbers of letters being sent.
Implemented an image classifier with PyTorch. The notebook or command line program was trained on a large dataset of flower images and is able to recognize over one hundred species and assing the correct label to a new image.
Using a real-life dataset representing German demographic information, I used unsupervised learning techniques to summarize and cluster unlabeled data into groups with similar properties. These groups were then compared to the customers of a mail-order company in order to determine which specific segments are represented in the customer base.