Experiments with distributed matrix factorization (MF). Presented at DataWorks Summit 2017, München.
It contains:
- Offline DSGD in Flink (algorithm by Gemulla et al. 2011)
- Online MF in Spark
- Online MF in Flink
- Combination of online and offline MF in Flink
Examples are to be reviewed by third-party and soon to be added.