Projects


MS-Project: Improving Monitoring of IT Systems using Machine Learning

The project investigated the feasability of using LSTM and autoencoder technology to predict failures in MSSQL databases before they became serious issues that resulted in potentially catastrofic system failures.

Read more

BS-Project: DeCP-Live

DeCP is a distributed search engine for web-scale image collections the sacrifices response-time for high throughput by batcing hundres or thousands of queries in a single query. It has be implemented to run on Apache Spark and has been evaluated with up to 40 billion high-dimensional features running on over 50 machines (800 virtual cores).

Read more

PhD-Project: Distributed extended Cluster Pruning (DeCP)

DeCP is a distributed search engine for web-scale image collections the sacrifices response-time for high throughput by batching hundreds or thousands of queries in a single query. It has be implemented to run on Apache Spark and has been evaluated with up to 40 billion high-dimensional features running on over 50 machines (800 virtual cores).

Read more