Mathematics Post Ostrom Colloquium: “Geometric Methods in Image Processing, Networks, and Machine Learning”
4:10 pm Neill 5W
Dr. Andrea L. Bertozzi
Geometric methods have revolutionized the field of image processing and image analysis. I will review some of these classical methods including image snakes, total variation minimization, image segmentation methods based on curve minimization, diffuse interface methods, and state of the art fast computational algorithms that include ideas from compressive sensing. Recently some of these concepts have shown promise for problems in high dimensional data analysis and machine learning on graphs. I will give an overview of some of the classical results from imaging and ongoing work in high dimensional and network data.