Coursework projects for ETH's Data Mining class in Fall 2017. [0]
The project were implemented on top of a provided, simplistic map-reduce framework in an inherently distributed fashion. In addition, some algorithms have been designed as online algorithms in order to handle larger quantities of data with very restricted resources.
The project topics revolved around:
- Locality sensitive hashing
- Stochastic gradient descent
- Clustering
- LinUCB for bandits