Ananth Mahadevan

Machine Learning PhD Student

University of Helsinki

Biography

I am a PhD student in the Department of Computer Science at the University of Helsinki. I am supervised by Associate Professor Michael Mathioudakis from the Algorithmic Data Science group. My PhD thesis is titled “Scaling and Maintaining Machine Learning Pipelines”.

Interests

Systems for Machine Learning
Machine Unlearning
Graphs and Networks

Education

MSc (Technology) in Machine Learning, Data Science and Artificial Intelligence, 2020
Aalto University, Finland
BEng in Computer Science & Engineering, 2018
R.V. College of Engineering, India

Publications

Ananth Mahadevan, Michael Mathioudakis (2024). Cost-aware Retraining for Machine Learning. KBS.

PDF Cite Poster DOI Code Arvix

Ananth Mahadevan, Michael Mathioudakis, Eetu Mäkelä, Mikko Tolonen (2024). Optimizing a Data Science System for Text Reuse Analysis. Arxiv.

PDF Cite Project Code Arxiv

Yann Ryan, Ananth Mahadevan, Mikko Tolonen (2023). A Comparative text similarity analysis of the works of Bernard Mandeville. Digital Enlightenment Studies.

PDF Cite Project DOI

David Rosson, Eetu Mäkelä, Ville Vaara, Ananth Mahadevan, Yann Ryan, Mikko Tolonen (2023). Reception Reader: Exploring Text Reuse in Early Modern British Publications. JOHD.

Cite Project DOI Demo Arxiv

Ananth Mahadevan, Arpit Merchant, Yanhao Wang, Michael Mathioudakis (2022). Robustness of Sketched Linear Classifiers to Adversarial Attacks. CIKM.

PDF Cite Code Poster DOI

Arpit Merchant, Ananth Mahadevan, Michael Mathioudakis (2022). Scalably Using Node Attributes and Graph Structure for Node Classification. Entropy.

PDF Cite Code DOI

Ananth Mahadevan, Michael Mathioudakis (2022). Certifiable Unlearning Pipelines for Logistic Regression: An Experimental Study. Machine Learning and Knowledge Extraction.

PDF Cite Code DOI Arvix Poster

Bruno Ordozgiti, Ananth Mahadevan, Antonis Matakos, Aristides Gionis (2022). Provable randomized rounding for minimum-similarity diversification. In DAMI.

PDF Cite Code DOI

Projects

HPC-HD: High Performance Computing for the Detection and Analysis of Historical Discourses

Using HPC to detect discourses from large historical corpora of the 18th century