National Research Projects

Dapprox

Swiss National Science Foundation 2017–2022

Dependency-aware Approximate Analytics and Processing Platform


Dapprox is a set of methods and software tools for fast and approximate analyses of resource usage series in real time. The goal is to predict potential anomalies (and propose solutions) by simultaneously taking into account accuracy requirements, maximum delays and available resources.

WiBD

Swiss National Science Foundation  2018–2021

Women in Big Data
 

The mission of WiBD is to organize activities to promote women in Big Data research with a dual focus on social and technical sciences. W e advocate the importance of an open and broad discussion on building, advancing, and maintaining careers for women in Big Data fields.

LoadOpt

Swiss National Science Foundation 2013–2017

Workload Characterization and Optimization for Multicore Systems
 

Multicore architectures have become the standard, both for general-purpose PCs and for high-end servers. Such platforms provide an environment that is ideally suited for consolidating multiple applications within the same machine to take full advantage of the available parallelism. 

GENiC

European FP7 ICT project 2013–2016

Globally Optimized Energy Efficient Data Centres 

The GENiC platform provides a set of services that are loosely coupled on the GENiC controller cluster to support energy optimisation. These provide functionalities to handle data centre (DC) model mapping, sensor telemetry (airflow, temperature and power), workload actuation, fault detection and diagnosis and historical data storage.

 
Research Team

Current TU Delft PhD Supervision


T. Younesian
04/2019
–present
TU Delft, The Netherlands
Project: Active learning for dirty label classification

M. Ghassi
02/2019
–present
TU Delft, The Netherlands
Project: Robust machine learning systems

 

 

 

Affiliated Postdoctoral Fellows at IBM

N. Ramashan
03/2017
12/2017
Project: Cloud-based security service

R. Birke
01/2012
06/2017
Project: Datacenter discovery and cognitive threat intelligence platform


PhD Research Project Co-advising at IBM


H. Zhang
09/201704/2018
Co-adviser of Dapprox
Swiss Federal Institute of Technology (ETH), Switzerland
Project: Distributed lower-precision machine learning

 

H. Sun
01/2016
11/2016
Co-adviser of LoadOpt
University of Lugano, Switzerland
Project: Trading-off accuracy and latency on big data processing platforms

 

A. Rosa
10/2013
04/2016

Co-adviser of LoadOpt
University of Lugano, Switzerland
Project: Characterizing and predicting failures of big data applications


PhD Intern Mentors at IBM


S. Cerf
07
11/2018
University of Grenoble, France
Project: Robust learning against noisy data

 

N. Ramashan
08
12/2016
KTH Royal Institute of Technology, Sweden
Project: Optimizing tail latency for in-memory data store

 

J. Xue
08
12/2015
College of William and Mary, USA
Project: Proactive datacenter ticket management using approximate time series prediction

 

S. Spicuglia
03/2014
08/2015
University of Lugano, Switzerland
Project: Resource and overload management for big data platforms

 

C. Wang
06
10/2014
Pennsylvania State University, USA
Project: Pricing design for co-located cloud tenants

 

D. Cavada
09/2013
08/2014
Bogazici University, Turkey
Project: Energy-aware scheduling design for datacenter (became her thesis dissertation)

 

A. Podzimek
08/2012
04/2013
Charles University, Czech Republic
Project: Partial load benchmarking (became his thesis dissertation)

 

D. Ansaloni
09/2010
04/2011
University of Lugano, Switzerland
Project: Consolidating workloads on multicore systems

 

Master Students Co-advising at IBM


J. Vallone
09/2014
06/2015
Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland
Master thesis: Contention detection by throttling: a black-box on-line approach

 

Y. Yin
02
07/2014
Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland
Master thesis: Optimizing energy, locality and priority in a MapReduce cluster

 
Past IBM Research Projects

Cognitive Appliances

2018

Designed active learning schemes for automatic information retrieval from PDF documents and generation of knowledge graphs.

Cognitive Threats Insights Platform

2016–2017

Developed a unified and resilient platform and leveraged big data analysis to detect potential threats for large-scale distributed systems.

Key Protect

2017–2018

Optimized and accelerated cryptographic key management for IBM Cloud service.

Data Center Discovery

2013–2015

Led a team to conduct data analytics on performance logs of large-scale data centers, build learning models, and develop ticket-management solutions.