In Preparation

Aftab: Benchmarking CNN Feature Extractors in Deep Reinforcement Learning

Authors: Shieenavaz, Zareshahraki, Nanni
Supervisor: Dr. Loris Nanni
Status: In preparation

Thesis

Graph-meta: Improving Metagenomic Classification with Graph-based Learning

Author: Shabnam Zareshahraki
Advisor: Dr. Matteo Comin
Institution: Università degli Studi di Padova
Completed: April 2026
GitHub | Paper

Master’s thesis proposing a hybrid GNN-based architecture for taxonomic classification of metagenomic sequences. The work combines deep learning with graph-structured representations of genomic assembly graphs, outperforming state-of-the-art sequence-based classifiers on real-world multi-omic benchmarks.

For collaboration or preprint requests, please get in touch