Non-Parametric Mixture Model Applied to Image Segmentation

• 
  Non-Parametric Mixture Model Applied to Image Segmentation
  Marilyn Vazquez
  A central problem in genomics is how to determine whether a particular mutation will be harmful to humans. In this task traditional mathematical modeling has proven less successful than machine learning, which pragmatically tries to make predictions without regard for biological reasonableness or model elegance. In this talk I'll discuss my transition from applied mathematics to data science, and how I've used machine learning to help pinpoint the causes of genetic disease.