Skip to content

Integrative Scalable Computing Laboratory

A research group at the Department of Information Technology, Uppsala Universtity.

  • Home
  • Projects
  • People
  • Publications
  • Teaching
  • Software
  • Recruitment
  • About us
  • Toggle search form

MSc thesis opportunities in privacy-preserving Machine Learning

Posted on November 10, 2019November 10, 2019 By admin

We have opportunities for a number of MSc thesis students to work with the group in the spring semester 2020.

Artificial intelligence is rapidly transforming our society. Machine learning models will be in every digital system we use, and it is imperative that we protect the integrity of data owners. In this project we work on training schemes, scalable implementations, and applications of Federated Learning – a recent approach to training ML models while keeping input data privacy of data owners.

Federated Machine Learning 

Federated machine learning has recently attracted a lot of attention both in industry and academia. Simply speaking, training proceeds by model updates on private data nodes, then weights are averaged by a server forming a  global model (schematic figure inline). While simple in concept, care needs to be taken to balance local model training with global synchronization to avoid poor convergence and to minimize communication rounds. FedML differs from standard distributed learning/optimization in that data cannot be assumed to be balanced across nodes, data may not be i.i.d., and we cannot assume consistent node uptime nor low-latency high-throughput networking between nodes. During 2017 and 2018, Google Research presented an approach to FedML based on TensorFlow targeting mobile devices [2,3]. Other prominent efforts include the open source project  OpenMinded (https://www.openmined.org/) and the latest API extension of Tensorflow federated [4]. Intel in collaboration with the University of Pennsylvania recently demonstrated a real-world case for FedML based on biomedical imaging [5]. Machine learning models that has been demonstrated in the FedML case include CNNs, LSTMs and conformal predictors [6]. In our group we are currently working on various aspects of FedML such as new federated ensemble methods and schemes to measure individual member contributions in a scalable fashion.

Potential thesis topics 

We have opportunities for MSc thesis students in a number of areas in privacy-preserving learning, such as: 

  1. Performance evaluation and optimization of federated learning algorithms for new application areas and/or models.      
  2. Development of new FedML schemes. 
  3. Development of scalable computing backends.  
  4. Decentralized implementations to enable FedML without a trusted-third party.
  5. Privacy-enhancing techniques such as differential privacy and secure multiparty computation. 

Research environment

The work will be conducted as part of the research group Integrative Scalable Computing Laboratory. ISCL is an interdisciplinary team working on the interface of scientific computing, machine learning and distributed systems. The group runs a number of eScience projects with funding from eSSENCE, SSF, VR and NIH. The MSc student will get the opportunity to participate in the work of the group during the semester the thesis is written, gaining insight into the academic work culture.  

Contact

Reach out to Andreas Hellander or Salman Toor to discuss opportunities:

Andreas: andreas.hellander@it.uu.se

Salman:  salman.toor@it.uu.se 

References

  1. Feng X., Qing, K., Meyer CH. and Chen Q., Deep convolutional neural  network for segmentation of thoracic organ-at-risk using cropped 3D images, Med. Phys., 46(5), 2019.  
  2. Konečný J,, Brendan McMahan H., X. Yu F., Richtárik P.,, Theertha Suresh A., Bacon D., Federated Learning: Strategies for imporving communication efficiency, ArXiv 1610.05492, 2016.
  3. K. Bonawitz et al., Towards Federated Learning at Scale: System Design,  ArXiv 1902.01046, 2019.
  4. Tensorflow federated, https://www.tensorflow.org/federated 
  5. Sheller MJ, Reina GA, Edwards B, Martin J, Bakas S., Multi-institutional Deep Learning Modeling Without Sharing Patient Data: A Feasibility Study on Brain Tumor Segmentation,, Lecture Notes in Computer Science book series (Volume 11383). 2019.
  6. Gauraha, N. and Spjuth, O. Synergy Conformal Prediction DiVA preprint. 360504 (2018). URL: urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-360504. 
  7. How to Backdoor Federated Learning, E. Bagdasaryan, A. Veit, Y. Hua, D. Estrin, V. Shmatikov, ArXiv, 1807.00459, 2017


Applied Cloud Computing, FedML, News, Open Positions

Post navigation

Previous Post: Challenges
Next Post: Addi Ait-Mlouk joins the lab as postdoc focusing on FedML

More Related Articles

“Ten simple rules” for establishing a national scale OpenStack cloud e-infrastructure for science Applied Cloud Computing
HASTE is granted 29 MSEK funding from SSF News
Trusted Execution Environment (TEE) For Federated Learning Data-Intensive Computing
Perspective: Stochastic Algorithms for Chemical Kinetics News
Fredrik Wrede presented at MLSB 2017 News
MSc projects available in multicellular systems biology mscprojects

Data-and simulation-driven life science. Much of our work in eScience and applied ML has applications in life science, and in Systems Biology in particular. We aim to enable data-and simulation-driven scientific discovery.

HASTE - a cloud native framework for intelligent processing of image streams: http://haste.research.it.uu.se/

Follow us on twitter

Andreas HellanderFollow

Andreas Hellander
Retweet on TwitterAndreas Hellander Retweeted
yannik_schaelteYannik Schälte@yannik_schaelte·
9 Jun

🎒Summer school: 𝐈𝐧𝐯𝐞𝐫𝐬𝐞 𝐩𝐫𝐨𝐛𝐥𝐞𝐦𝐬 𝐟𝐨𝐫 𝐦𝐮𝐥𝐭𝐢-𝐬𝐜𝐚𝐥𝐞 𝐦𝐨𝐝𝐞𝐥𝐬

💬With: Linda Petzold, Christiane Fuchs, @dennisprangle, @StefanEngblom, @A_Hellander

⏰August 22-26
📌@HCM_Bonn

✏️Details+register: http://www.hcm.uni-bonn.de/events/eventpages/hausdorff-school/hausdorff-schools-2022/inverse-2022/

Reply on Twitter 1534886971478261770Retweet on Twitter 15348869714782617709Like on Twitter 153488697147826177019Twitter 1534886971478261770
A_HellanderAndreas Hellander@A_Hellander·
25 May

Really enjoyed presenting our work on federated learning with FEDn at CCGRID22 last week. Some takeaways from the talk (1/3):

1. Algorithm development must have real-world scalability in mind and there is a risk in missing this aspect if only considering simulation of FL.

Reply on Twitter 1529360630087557120Retweet on Twitter 1529360630087557120Like on Twitter 15293606300875571201Twitter 1529360630087557120
A_HellanderAndreas Hellander@A_Hellander·
25 May

Such a beautiful paper put together by @adameykolab, and one of the most fun applications of fluid mechanics modeling I have seen in recent years @AnassBouchnita @MurtazoNazarov. Thanks for the collaboration!

Igor Adameyko@adameykolab

After a long struggle, failed revision in Science, loads of happiness and pain, our paper on surface-associated water streams integrating polyps into a coral colony, is out in Current Biology. https://www.cell.com/current-biology/fulltext/S0960-9822(22)00672-8?fbclid=IwAR0j5TqfenF0_tp1hDam43K6jdfWS9iw16OzGVHZsVZ89eDe0Yq8Aa-ykuY#%20 1/10

Reply on Twitter 1529359602516738049Retweet on Twitter 1529359602516738049Like on Twitter 15293596025167380493Twitter 1529359602516738049
Retweet on TwitterAndreas Hellander Retweeted
adameykolabIgor Adameyko@adameykolab·
14 May

After a long struggle, failed revision in Science, loads of happiness and pain, our paper on surface-associated water streams integrating polyps into a coral colony, is out in Current Biology. https://www.cell.com/current-biology/fulltext/S0960-9822(22)00672-8?fbclid=IwAR0j5TqfenF0_tp1hDam43K6jdfWS9iw16OzGVHZsVZ89eDe0Yq8Aa-ykuY#%20 1/10

Reply on Twitter 1525473700971175936Retweet on Twitter 152547370097117593631Like on Twitter 1525473700971175936158Twitter 1525473700971175936
A_HellanderAndreas Hellander@A_Hellander·
11 May

Are you using StochSS? Please help us gather insights into what is working well and what can be improved by filling in this short survey https://forms.gle/mEqfASuUd3MDWuPS9

@LindaPetzold @briandrawert @mhucka

Reply on Twitter 1524477596930654214Retweet on Twitter 15244775969306542141Like on Twitter 15244775969306542141Twitter 1524477596930654214
Load More...

Decentralized AI, Federated Learning. One focus area of the group is development of methods and software to address decentralized and privacy-preserving AI. We are core contributors to the FEDn open source framework for scalable federated machine learning:

https://github.com/scaleoutsystems/fedn
Introduction to Federated Learning by Andreas Hellander
Join the discussion on Decentralized AI:

Scaleout Systems is a spin-out from ISCL on a mission to enable decentralized AI and federated learning to production.

https://www.scaleoutsystems.com/

Copyright © 2022 Integrative Scalable Computing Laboratory.

Powered by PressBook Blog WordPress theme