Thesis' proposals

This list is incomplete. If you are interested in my research topics (see here ), contact me for more information (here a few keywords describing my work: similarity search, graph mining, parallel algorithms, I/O efficient algorithms, fairness in machine learning,…)


Proposals for bachelor students (“Laurea Triennale”)

Thesis topics: similarity search, graph mining, parallel algorithms, I/O efficient algorithms, fairness in machine learning,…

If you are interested, write me!


Proposals for master students (“Laurea Magistrale”)

Algorithms for Tensor Processing Units

The goal of this master thesis is to develop and analyse algorithms for Tensor Processing Units (TPUs). A TPU is a computing unit optimised for deep learning, which features special hardware for dense matrix multiplication. The goal of this thesis is to show that TPU can be used for efficiently solving sparse problems. In this thesis, the student will study some recent parallel algorithms (see e.g., [1]) for sparse problems and adapt them to TPUs. The algorithms will then be tested on a NVIDIA Volta board [2]. The thesis will be carried out in collaboration with Dr. Flavio Vella from Free University of Bozen.

References:

  1. J. Kepner and J. Gilbert. Graph Algorithms in the Language of Linear Algebra. SIAM, 2011

  2. Nvidia TPU: https://devblogs.nvidia.com/programming-tensor-cores-cuda-9/