Research

2024

1. NeurIPS

   

   PaPaGei: Open Foundation Models for Optical Physiological Signals

   Arvind Pillai, Dimitris Spathis, Fahim Kawsar, and Mohammad Malekzadeh

   In NeurIPS Workshop on Time Series in the Age of Large Models, Dec 2024

   Bib HTML arXiv Code

   @inproceedings{das2024primus,
     title = {PaPaGei: Open Foundation Models for Optical Physiological Signals},
     author = {Pillai, Arvind and Spathis, Dimitris and Kawsar, Fahim and Malekzadeh, Mohammad},
     booktitle = {NeurIPS Workshop on Time Series in the Age of Large Models},
     year = {2024},
     month = dec,
   }

2. NeurIPS

   

   PRIMUS: Pretraining IMU Encoders with Multimodal and Self-Supervisied Learning

   Arnav M Das, Chi Ian Tang, Fahim Kawsar, and Mohammad Malekzadeh

   In NeurIPS Workshop on Time Series in the Age of Large Models, Dec 2024

   Bib HTML

   @inproceedings{das2024primut,
     title = {PRIMUS: Pretraining IMU Encoders with Multimodal and Self-Supervisied Learning},
     author = {Das, Arnav M and Tang, Chi Ian and Kawsar, Fahim and Malekzadeh, Mohammad},
     booktitle = {NeurIPS Workshop on Time Series in the Age of Large Models},
     year = {2024},
     month = dec,
   }

3. NeurIPS

   

   Analysing Softmax Entropy Minimization for Adaptating Multitask Models at Test-time

   Soumyajit Chatterjee, Abhirup Ghosh, Fahim Kawsar, and Mohammad Malekzadeh

   In NeurIPS Workshop on Fine-Tuning in Modern Machine Learning: Principles and Scalability, Dec 2024

   Bib HTML

   @inproceedings{das2024primuu,
     title = {Analysing Softmax Entropy Minimization for Adaptating Multitask Models at Test-time},
     author = {Chatterjee, Soumyajit and Ghosh, Abhirup and Kawsar, Fahim and Malekzadeh, Mohammad},
     booktitle = {NeurIPS Workshop on Fine-Tuning in Modern Machine Learning: Principles and Scalability},
     year = {2024},
     month = dec,
   }

4. ICLR

   

   SoundCollage: Automated Discovery of New Classes in Audio Datasets

   Ryuhaerang Choi, Soumyajit Chatterjee, Dimitris Spathis, Sung-Ju Lee, Fahim Kawsar, and Mohammad Malekzadeh

   In ICLR Workshop on Data-centric Machine Learning Research (DMLR), May 2024

   Bib HTML arXiv Code

   @inproceedings{das2024primuv,
     title = {SoundCollage: Automated Discovery of New Classes in Audio Datasets},
     author = {Choi, Ryuhaerang and Chatterjee, Soumyajit and Spathis, Dimitris and Lee, Sung-Ju and Kawsar, Fahim and Malekzadeh, Mohammad},
     booktitle = {ICLR Workshop on Data-centric Machine Learning Research (DMLR)},
     year = {2024},
     month = may,
   }

5. arXiv

   

   Enhancing Efficiency in Multidevice Federated Learning through Data Selection

   Fan Mo, Mohammad Malekzadeh, Soumyajit Chatterjee, Fahim Kawsar, and Akhil Mathur

   ArXiv Preprint, May 2024

   Bib arXiv Code

   @article{mo2024enhancing,
     title = {Enhancing Efficiency in Multidevice Federated Learning through Data Selection},
     author = {Mo, Fan and Malekzadeh, Mohammad and Chatterjee, Soumyajit and Kawsar, Fahim and Mathur, Akhil},
     journal = {ArXiv Preprint},
     year = {2024},
   }

6. arXiv

   

   Deep Unlearn: Benchmarking Machine Unlearning

   Xavier F. Cadet, Anastasia Borovykh, Mohammad Malekzadeh, Sara Ahmadi-Abhari, and Hamed Haddadi

   ArXiv Preprint, May 2024

   Bib HTML arXiv

   @article{cadet2024deep,
     title = {Deep Unlearn: Benchmarking Machine Unlearning},
     author = {Cadet, Xavier F. and Borovykh, Anastasia and Malekzadeh, Mohammad and Ahmadi-Abhari, Sara and Haddadi, Hamed},
     journal = {ArXiv Preprint},
     year = {2024},
   }

7. BMVC

   

   Vicious Classifiers: Assessing Inference-time Data Reconstruction Risk in Edge Computing

   Mohammad Malekzadeh, and Deniz Gunduz

   In BMVC 2024 workshop on Privacy, Fairness, Accountability and Transparency in Computer Vision, Nov 2024

   Bib arXiv Code

   @inproceedings{malekzadeh2024vicious,
     title = {Vicious Classifiers: Assessing Inference-time Data Reconstruction Risk in Edge Computing},
     author = {Malekzadeh, Mohammad and Gunduz, Deniz},
     booktitle = {BMVC 2024 workshop on Privacy, Fairness, Accountability and Transparency in Computer Vision},
     year = {2024},
     month = nov,
   }

8. WSDM

   

   CroSSL: Cross-modal Self-Supervised Learning for Time-series through Latent Masking

   Shohreh Deldari, Dimitrios Spathis, Mohammad Malekzadeh, Fahim Kawsar, Flora Salim, and Akhil Mathur

   In 17th ACM International Conference on Web Search and Data Mining (WSDM), Mar 2024

   Bib arXiv Code

   @inproceedings{deldari2024crossl,
     title = {CroSSL: Cross-modal Self-Supervised Learning for Time-series through Latent Masking},
     author = {Deldari, Shohreh and Spathis, Dimitrios and Malekzadeh, Mohammad and Kawsar, Fahim and Salim, Flora and Mathur, Akhil},
     booktitle = {17th ACM International Conference on Web Search and Data Mining (WSDM)},
     year = {2024},
     month = mar,
   }

9. HotMobile

   

   Salted Inference: Enhancing Privacy while Maintaining Efficiency of Split Inference in Mobile Computing

   Mohammad Malekzadeh, and Fahim Kawsar

   In 25th International Workshop on Mobile Computing Systems and Applications (HotMobile), Feb 2024

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2024salted,
     title = {Salted Inference: Enhancing Privacy while Maintaining Efficiency of Split Inference in Mobile Computing},
     author = {Malekzadeh, Mohammad and Kawsar, Fahim},
     booktitle = {25th International Workshop on Mobile Computing Systems and Applications (HotMobile)},
     year = {2024},
     month = feb,
   }

2023

1. ICLR

   

   Centaur: Federated Learning for Constrained Edge Devices

   Fan Mo, Mohammad Malekzadeh, Soumyajit Chatterjee, Fahim Kawsar, and Akhil Mathur

   In ICLR Workshop on Machine Learning for IoT: Datasets, Perception, and Understanding (ML4IoT), May 2023

   Bib HTML arXiv Code Extended Version

   @inproceedings{mo2023centaur,
     title = {Centaur: Federated Learning for Constrained Edge Devices},
     author = {Mo, Fan and Malekzadeh, Mohammad and Chatterjee, Soumyajit and Kawsar, Fahim and Mathur, Akhil},
     booktitle = {ICLR Workshop on Machine Learning for IoT: Datasets, Perception, and Understanding (ML4IoT)},
     year = {2023},
     month = may,
     crossref = {mo2024enhancing},
   }

2. arXiv

   

   Vicious Classifiers: Data Reconstruction Attack at Inference Time

   Mohammad Malekzadeh, and Deniz Gunduz

   ArXiv Preprint, Dec 2023

   [The extended version is available in the BMVC version above.]

   Bib arXiv Code Extended Version

   @article{malekzadeh2023vicious,
     title = {Vicious Classifiers: Data Reconstruction Attack at Inference Time},
     author = {Malekzadeh, Mohammad and Gunduz, Deniz},
     journal = {ArXiv Preprint},
     year = {2023},
     month = dec,
     note = {[The extended version is available in the BMVC version above.]},
     crossref = {malekzadeh2024vicious},
   }

3. ICML

   

   Latent Masking for Multimodal Self-supervised Learning in Health Timeseries

   Shohreh Deldari, Dimitrios Spathis, Mohammad Malekzadeh, Fahim Kawsar, Flora Salim, and Akhil Mathur

   In ICML Workshop on Machine Learning for Multimodal Healthcare Data (ML4MHD), Jul 2023

   Bib arXiv Code Extended Version

   @inproceedings{deldari2023latent,
     title = {Latent Masking for Multimodal Self-supervised Learning in Health Timeseries},
     author = {Deldari, Shohreh and Spathis, Dimitrios and Malekzadeh, Mohammad and Kawsar, Fahim and Salim, Flora and Mathur, Akhil},
     booktitle = {ICML Workshop on Machine Learning for Multimodal Healthcare Data (ML4MHD)},
     year = {2023},
     month = jul,
     crossref = {deldari2024crossl},
   }

2022

1. NSDI

   

   Re-architecting Traffic Analysis with Neural Network Interface Cards

   Giuseppe Siracusano, Salvator Galea, Davide Sanvit, Mohammad Malekzadeh, Paolo Costa, Gianni Antichi, Hamed Haddadi, and Roberto Bifulco

   In 19th USENIX Symposium on Networked Systems Design and Implementation (NSDI), Jul 2022

   Bib HTML

   @inproceedings{siracusano2022rearchitecting,
     title = {Re-architecting Traffic Analysis with Neural Network Interface Cards},
     author = {Siracusano, Giuseppe and Galea, Salvator and Sanvit, Davide and Malekzadeh, Mohammad and Costa, Paolo and Antichi, Gianni and Haddadi, Hamed and Bifulco, Roberto},
     booktitle = {19th USENIX Symposium on Networked Systems Design and Implementation (NSDI)},
     year = {2022},
   }

2. arXiv

   

   Quantifying and Localizing Usable Information Leakage from Neural Network Gradients

   Fan Mo, Anastasia Borovykh, Mohammad Malekzadeh, Soteris Demetriou, Deniz Gündüz, and Hamed Haddadi

   ArXiv Preprint, Jul 2022

   Bib arXiv

   @article{mo2021quantifying,
     title = {Quantifying and Localizing Usable Information Leakage from Neural Network Gradients},
     author = {Mo, Fan and Borovykh, Anastasia and Malekzadeh, Mohammad and Demetriou, Soteris and Gündüz, Deniz and Haddadi, Hamed},
     journal = {ArXiv Preprint},
     year = {2022},
     month = jul,
   }

2021

1. CCS

   

   Honest-but-Curious Nets: Sensitive Attributes of Private Inputs can be Secretly Coded into the Classifiers’ Outputs

   Mohammad Malekzadeh, Anastasia Borovykh, and Deniz Gündüz

   In ACM Conference on Computer and Communications Security (CCS), Nov 2021

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2021honest,
     title = {Honest-but-Curious Nets: Sensitive Attributes of Private Inputs can be Secretly Coded into the Classifiers' Outputs},
     author = {Malekzadeh, Mohammad and Borovykh, Anastasia and Gündüz, Deniz},
     booktitle = {ACM Conference on Computer and Communications Security (CCS)},
     year = {2021},
     month = nov,
   }

2. CCS

   

   Efficient Hyperparameter Optimization for Differentially Private Deep Learning

   Aman Priyanshu, Rakshit Naidu, Fatemehsadat Mireshghallah, and Mohammad Malekzadeh

   In CCS Workshop on Privacy Preserving Machine Learning (ACM CCS), Nov 2021

   Bib HTML arXiv

   @inproceedings{priyanshu2021efficient,
     title = {Efficient Hyperparameter Optimization for Differentially Private Deep Learning},
     author = {Priyanshu, Aman and Naidu, Rakshit and Mireshghallah, Fatemehsadat and Malekzadeh, Mohammad},
     booktitle = {CCS Workshop on Privacy Preserving Machine Learning (ACM CCS)},
     year = {2021},
     month = nov,
   }

3. IMWUT

   

   DANA: Dimension-Adaptive Neural Architecture for Multivariate Sensor Data

   Mohammad Malekzadeh, Richard G. Clegg, Andrea Cavallaro, and Hamed Haddadi

   ACM Journal on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT), Oct 2021

   [Presented at the ACM conference for Ubiquitous Computing (UbiComp 2021)]

   Bib HTML arXiv Code

   @article{malekzadeh2021dana,
     title = {DANA: Dimension-Adaptive Neural Architecture for Multivariate Sensor Data},
     author = {Malekzadeh, Mohammad and Clegg, Richard G. and Cavallaro, Andrea and Haddadi, Hamed},
     journal = {ACM Journal on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT)},
     year = {2021},
     month = oct,
     note = {[Presented at the ACM conference for Ubiquitous Computing (UbiComp 2021)]},
   }

4. ICLR

   

   Layer-wise characterization of latent information leakage in federated learning

   Fan Mo, Anastasia Borovykh, Mohammad Malekzadeh, Soteris Demetriou, and Hamed Haddadi

   In ICLR Workshop on Distributed and Private Machine Learning (DPML), May 2021

   [The extended version is available in the arXiv version above.]

   Bib HTML arXiv Extended Version

   @inproceedings{mo2021quantifyinh,
     title = {Layer-wise characterization of latent information leakage in federated learning},
     author = {Mo, Fan and Borovykh, Anastasia and Malekzadeh, Mohammad and Demetriou, Soteris and Haddadi, Hamed},
     booktitle = {ICLR Workshop on Distributed and Private Machine Learning (DPML)},
     year = {2021},
     month = may,
     note = {[The extended version is available in the arXiv version above.]},
     crossref = {mo2021quantifying},
   }

5. AAAI

   

   Dopamine: Differentially Private Federated Learning on Medical Data

   Mohammad Malekzadeh, Burak Hasircioglu, Nitish Mital, Kunal Katarya, Mehmet Emre Ozfatura, and Deniz Gündüz

   In The Second AAAI Workshop on Privacy-Preserving Artificial Intelligence (PPAI), Feb 2021

   [Selected as the best submission to the ITU AI/ML in 5G Challenge, sub-challenge Privacy Preserving AI/ML in 5G networks for healthcare applications.]

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2021dopamine,
     title = {Dopamine: Differentially Private Federated Learning on Medical Data},
     author = {Malekzadeh, Mohammad and Hasircioglu, Burak and Mital, Nitish and Katarya, Kunal and Ozfatura, Mehmet Emre and Gündüz, Deniz},
     booktitle = {The Second AAAI Workshop on Privacy-Preserving Artificial Intelligence (PPAI)},
     year = {2021},
     month = feb,
     note = {[Selected as the best submission to the ITU AI/ML in 5G Challenge, sub-challenge Privacy Preserving AI/ML in 5G networks for healthcare applications.]},
   }

2020

1. MLSys

   

   Privacy-Preserving Bandits

   Mohammad Malekzadeh, Dimitrios Athanasakis, Hamed Haddadi, and Benjamin Livshits

   In Conference on Machine Learning and Systems (MLSys), Austin, USA, Mar 2020

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2020privacy_bandits,
     title = {Privacy-Preserving Bandits},
     author = {Malekzadeh, Mohammad and Athanasakis, Dimitrios and Haddadi, Hamed and Livshits, Benjamin},
     booktitle = {Conference on Machine Learning and Systems (MLSys)},
     year = {2020},
     month = mar,
     location = {Austin, USA},
   }

2. RSOS

   

   Modeling and Forecasting Art Movements with CGANs

   Edoardo Lisi, Mohammad Malekzadeh, Hamed Haddadi, F. Din-Houn Lau, and Seth Flaxman

   Royal Society Open Science Journal (RSOS), Feb 2020

   Bib HTML arXiv Code

   @article{lisi2020modeling,
     title = {Modeling and Forecasting Art Movements with CGANs},
     author = {Lisi, Edoardo and Malekzadeh, Mohammad and Haddadi, Hamed and Lau, F. Din-Houn and Flaxman, Seth},
     journal = {Royal Society Open Science Journal (RSOS)},
     year = {2020},
     month = feb,
     doi = {10.1098/rsos.191569},
   }

3. arXiv

   

   Running Neural Networks on the NIC

   Giuseppe Siracusano, Salvator Galea, Davide Sanvit, Mohammad Malekzadeh, Paolo Costa, Gianni Antichi, Hamed Haddadi, and Roberto Bifulco

   ArXiv Preprint, Sep 2020

   Bib arXiv

   @article{siracusano2022rearchitectinh,
     title = {Running Neural Networks on the NIC},
     author = {Siracusano, Giuseppe and Galea, Salvator and Sanvit, Davide and Malekzadeh, Mohammad and Costa, Paolo and Antichi, Gianni and Haddadi, Hamed and Bifulco, Roberto},
     journal = {ArXiv Preprint},
     year = {2020},
     month = sep,
   }

2019

1. PMC

   

   Privacy and Utility Preserving Sensor-Data Transformations

   Mohammad Malekzadeh, Richard G. Clegg, Andrea Cavallaro, and Hamed Haddadi

   Pervasive and Mobile Computing Journal, Elsevier, Nov 2019

   Bib HTML arXiv Code

   @article{malekzadeh2020privacy,
     title = {Privacy and Utility Preserving Sensor-Data Transformations},
     author = {Malekzadeh, Mohammad and Clegg, Richard G. and Cavallaro, Andrea and Haddadi, Hamed},
     journal = {Pervasive and Mobile Computing Journal, Elsevier},
     year = {2019},
     month = nov,
     doi = {10.1016/j.pmcj.2020.101132},
   }

2. IoTDI

   

   Mobile Sensor Data Anonymization

   Mohammad Malekzadeh, Richard G. Clegg, Andrea Cavallaro, and Hamed Haddadi

   In ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI), Montreal, Canada, Apr 2019

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2019mobile,
     title = {Mobile Sensor Data Anonymization},
     author = {Malekzadeh, Mohammad and Clegg, Richard G. and Cavallaro, Andrea and Haddadi, Hamed},
     booktitle = {ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI)},
     year = {2019},
     month = apr,
     location = {Montreal, Canada},
     doi = {10.1145/3302505.3310068},
   }

2018

1. EuroSys

   

   Protecting Sensory Data against Sensitive Inferences

   Mohammad Malekzadeh, Richard G. Clegg, Andrea Cavallaro, and Hamed Haddadi

   In EuroSys Workshop on Privacy by Design in Distributed Systems (W-P2DS), Porto, Portugal, Apr 2018

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2018protecting,
     title = {Protecting Sensory Data against Sensitive Inferences},
     author = {Malekzadeh, Mohammad and Clegg, Richard G. and Cavallaro, Andrea and Haddadi, Hamed},
     booktitle = {EuroSys Workshop on Privacy by Design in Distributed Systems (W-P2DS)},
     year = {2018},
     month = apr,
     location = {Porto, Portugal},
     doi = {10.1145/3195258.3195260},
   }

2. IoTDI

   

   Replacement AutoEncoder: A Privacy-Preserving Algorithm for Sensory Data Analysis

   Mohammad Malekzadeh, Richard G. Clegg, and Hamed Haddadi

   In ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI), Florida, USA, Apr 2018

   Bib HTML arXiv Code

   @inproceedings{malekzadeh2018replacement,
     title = {Replacement AutoEncoder: A Privacy-Preserving Algorithm for Sensory Data Analysis},
     author = {Malekzadeh, Mohammad and Clegg, Richard G. and Haddadi, Hamed},
     booktitle = {ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI)},
     year = {2018},
     month = apr,
     location = {Florida, USA},
     doi = {10.1109/IoTDI.2018.00022},
   }

3. Alan-Turing-Institute

   

   Technical Report: Fairness in Algorithmic Decision Making

   Paul-Marie Carfantan, Omar Costilla-Reyes, Delia Fuhrmann, Jonas Glesaaen, Qi He, Andreas Kirsch, Julie Lee, Mohammad Malekzadeh, Esben Sø rig, Caryn Tan, Emily Turner, and Vinh Quang

   , London, UK, Apr 2018

   Bib HTML

   @techreport{carfantan2018fairness,
     title = {Technical Report: Fairness in Algorithmic Decision Making},
     author = {Carfantan, Paul-Marie and Costilla-Reyes, Omar and Fuhrmann, Delia and Glesaaen, Jonas and He, Qi and Kirsch, Andreas and Lee, Julie and Malekzadeh, Mohammad and rig, Esben S\o and Tan, Caryn and Turner, Emily and Quang, Vinh},
     institution = {Data Study Group team. Data Study Group Final Report: Accenture. Zenodo},
     year = {2018},
     month = apr,
     location = {London, UK},
     doi = {10.5281/zenodo.2557795},
   }

2017

1. IJACSA

   

   Gamified Incentives: A Badge Recommendation Model to Improve User Engagement in Social Networking Websites

   Reza Gharibi, and Mohammad Malekzadeh

   International Journal of Advanced Computer Science and Applications (IJACSA), Nov 2017

   Bib HTML Code

   @article{gharibi2017gamified,
     title = {Gamified Incentives: A Badge Recommendation Model to Improve User Engagement in Social Networking Websites},
     author = {Gharibi, Reza and Malekzadeh, Mohammad},
     journal = {International Journal of Advanced Computer Science and Applications (IJACSA)},
     year = {2017},
     month = nov,
   }

2014

1. ASONAM

   

   A Multi-Generational Social Learning Model: the Effect of Information Cascade on Aggregate Welfare

   Marziyeh Barghandan, Mohammad Malekzadeh, Atefeh Safdel, and Iren Mazloomzadeh

   In The 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), Aug 2014

   Bib HTML PDF

   @inproceedings{barghandan2014multi,
     title = {A Multi-Generational Social Learning Model: the Effect of Information Cascade on Aggregate Welfare},
     author = {Barghandan, Marziyeh and Malekzadeh, Mohammad and Safdel, Atefeh and Mazloomzadeh, Iren},
     booktitle = {The 2014 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)},
     year = {2014},
     month = aug,
     address = {Beijing, China},
   }

2011

1. SNA-KDD

   

   Social Balance and Signed Network Formation Games

   Mohammad Malekzadeh, MohammadAmin Fazli, Pooya Jalaly Khalilabadi, Hamid R. Rabiee, and MohammadAli Safari

   In The 5th SNA-KDD Workshop (SNA-KDD), Aug 2011

   Bib PDF

   @inproceedings{malekzadeh2011social,
     title = {Social Balance and Signed Network Formation Games},
     author = {Malekzadeh, Mohammad and Fazli, MohammadAmin and Khalilabadi, Pooya Jalaly and Rabiee, Hamid R. and Safari, MohammadAli},
     booktitle = {The 5th SNA-KDD Workshop (SNA-KDD)},
     year = {2011},
     month = aug,
     address = {San Diego, USA},
   }

2008

1. CSICC

   

   Persian CAPTCHA System to Prevent Automatic Subscribing of Software Robots in Web Pages

   Mohammad Malekzadeh, and Mehdy Bohlool

   In The 13th National CSI Computer Conference (CSICC), Mar 2008

   [This paper is published in Persian]

   Bib PDF

   @inproceedings{malekzadeh2008persian,
     title = {Persian {CAPTCHA} System to Prevent Automatic Subscribing of Software Robots in Web Pages},
     author = {Malekzadeh, Mohammad and Bohlool, Mehdy},
     booktitle = {The 13th National CSI Computer Conference (CSICC)},
     year = {2008},
     month = mar,
     address = {Kish Island, Iran},
     note = {[This paper is published in Persian]},
   }