Noisy Label Learning: How to Handle the Dirty Secret of AI
Dirty data abounds in observational studies. No one designed the data collection process to apply AI methods, but everyone still expects you to make something out of it. Perhaps it’s too expensive to collect more, or it will take years to gather it. Come, commiserate about your data woes, and learn some of the latest techniques to spin that straw data into gold.
Taylor Brown is a Principal Data Scientist for CoreLogic, where he is pioneering the use of Deep Learning and imagery analytics to transform the real estate business. Prior to CoreLogic, Taylor was a Research Data Scientist for a boutique consulting firm in Virginia. Taylor has 15 years of experience developing and executing software and AI solutions for aerospace, education, and real estate industries. Taylor holds a M.S. in Computer Science from Columbia University.