Flink-online-machine-learning

Distributed Online Machine Learning Algorithms in Flink

Collaborators:

Zhanwang Chen, zhanwangchen
Oleksandr Chumak, xelabit Changbin Lu, fine1x

Task summary

The main task of the project was to implement a set of different machine learning algorithms in Flink's Streaming Environment using the Parameter Server and pipeline paradigms and evaluating the results afterwards.

algorithms consists of:

1. Online Support Vector Machine (OSVM)
2. Online Ridge Regression (ORR)
3. Lasso w/ Shrinkage via Limit of Gibbs Sampling (SLOG)
4. Asynchronous Stochastic Variational Inference (ASVI)
5. Growing Gaussian Mixture Model (GGMM)

Links of the project:

• Report
• Sources