){kind=avatar}
Graduate Student in Artificial Intelligence and Robotics
at La Sapienza
arturbackdeluca [at] gmail [dot] com
I am a master's student in Artificial Intelligence and Robotics at La Sapienza, in Rome.
I am currently looking for research opportunities mainly involved with the investigation of training dynamics of neural networks and their theoretical guarantees.
Please feel free to reach me via any of the digital mediums here available. Due the pandemic, I continued my studies from Brazil, and now I’m based in Toronto. My favorite activities are discovering new genres of music and running, usually at the same time 😃
Interests
- Statistical Learning Theory
- Optimization
- Graph Neural Networks
Education
-
M.Sc. in Artificial Intelligence and Robotics, 2019 -
La Sapienza University of Rome
-
Exchange Program, 2017 - 2018
RWTH Aachen
-
B.Sc. in Mechanical Engineering, 2013 - 2019
Federal University of Santa Catarina (UFSC)
Posts
A glimpse over capacity control in neural networks
By this point in time, there’s no need to emphasize how important deep learning has become. Yet, despite paving the way for some of the boldest and most complicated projects of this decade, much of deep learning’s theoretical guarantees remain unexplained.
Unsupervised pretraining
In the early development of neural networks, the initialization of parameters was performed by sampling a uniform distribution bounded to a predetermined interval. In this process, the weights and biases produced solutions that would generally lie far from an optimum set of parameters, requiring many training iterations to readjust.
Projects
ECG GANs
Classifying arrhytmias in ECGs using Generative Adversarial Networks (GANs) and spectral analysis
Testing the Lottery Ticket Hypothesis
Series of experiments replicating Frankle and Carbin’s Lottery Ticket Hypothesis
Landscapeviz
Visualizing the loss landscape of fully connected neural networks using tensorflow
PSOpt - Particle swarm optimizer
A python package for discrete optimization using particle swarms
Soft Actor-Critic in MXNet
Implementation and evaluation of the Soft-Actor critic algorithm using MXNet and OpenAI Gym
