Stanford MS&E 326 – Free and incentivized exploration in online learningClass description – Spring 2019Online learning problems typically require a balancing act between exploration and exploitation. Recent research has considered a range of scenarios where the learning algorithm essentially learns “for free”; these are situations where greedy algorithms work well. Another line of literature has investigated how a platform can incentivize exploration by paying agents to explore. This doctoral seminar will survey recent literature on free and incentivized exploration and related topics. A detailed syllabus will be distributed in the first lecture. Prerequisites: Strong background in optimization and stochastic analysis. Intended for doctoral students pursuing research in related areas. LogisticsClass times and locations:
|