Curriculum vitae

Profile

I am a Machine Learning researcher with a background in probabilistic models, reinforcement learning, computer vision and logic programming.

My main expertise includes:
Monte-Carlo methods, Bayesian inference, Reinforcement Learning, MDPs, Relational Learning, Logic programming, Computer vision, Deep Learning.

I am particularly interested in Reinforcement Learning, Deep Learning and their combination.

Work experience

04/09/2017–ongoing

Senior Machine learning Engineer

Prophesee, Paris, company that produces neuromorphic (event-based) cameras.

My role is the development of machine learning models for the camera such as object detection and optical flow.

19/02/2010–31/07/2010

Software developer

Dreamslair Entertainment s.r.l., Valenzano (Italy)

Analysis and development of browser-games (server-side and client-side).

01/08/2007–31/07/2008

Research project

UNIVERSITA' DEGLI STUDI DI MILANO - BICOCCA , Department of Informatics, Systems and Communication, Milano (Italy)

collaboration in "Neuroweb" project (Integration and sharing of information and knowledge in neurology and neurosciences)

04/09/2007–31/07/2008

Collaboration

@ITIM (Italian Association of Telemedicine and Medical Informatics), Desio (Italy) Organizational and didactic collaboration in meetings and seminaries:

"Grid technologies in biomedicine and healthcare" at "Mediterranean Meeting of Telemedicine"
Seminar on Data Analysis and Statistic Software in Medicine. Main topics: neural network MLP and SOM
Design and Development of a web portal to manage health information. Main technologies: PHP and SSL.

Education

31/08/2016–31/02/2017

Postdoctoral researcher

KU Leuven, Leuven (Belgium)

Involved in EU ReGROUND project, whose goal is to perform symbol grounding to link human language, internal robot belief and perception for robotic tasks. Research:

RelNet: Relational Neural Network to perform relational learning with the power of neural networks
Relational symbol grounding: learn the meaning of relations extracted from sentences

Teaching (exercise sessions) “Machine Learning and Inductive Inference”

01/03/2011–30/08/2016

Doctor of Engineering Science (PhD) in Computer Science

KU Leuven, Leuven (Belgium)

PhD thesis: Hybrid Probabilistic Logic Programming

Main topics: Probabilistic programming, Logic programming, Bayesian inference, Monte-Carlo methods, Particle filtering, probabilistic planning (MDPs), object tracking, Robotics.

Involved in EU FP7 First-MM project, whose goal is to develop autonomous mobile manipulation robots that can perform complex manipulation and transportation tasks.

Obtained IWT (government agency for Innovation by Science and Technology) scholarship from 2012.

Teaching and supervision

Teaching (exercise sessions) and projects correction for “Machine Learning and Inductive Inference”
Master student thesis supervision (main topics: relational learning, Robotics, learning by demonstration)

01/10/2010-20/12/2010

Intership

Institute of Cognitive Sciences and Technologies − CNR (Rome - Italy)

Internship in EU FP7 Humanobs project, whose goal is to learn socio-communicative skills by observing and imitating people.

2005–2009

Master degree in Computer Systems Engineering, specialization in intelligent systems

Politecnico di Bari, Facoltà di Ingegneria (Polytechnic of Bari, Faculty of Engineering), Bari (Italy)

Final Mark: 110/110 cum laude

Master thesis: "Boolean Games and Description Logics for Multi-attribute Automatic Negotiation"

Main topics: mathematics, computer science, electronics, control, physics.

2002–2005

Bachelor Degree in Computer Systems Engineering

Politecnico di Bari, Facoltà di Ingegneria (Polytechnic of Bari, Faculty of Engineering), Bari (Italy)

Final Mark: 110/110 cum laude

Bachelor thesis: "Stereo-Matching Techniques Optimization Using Evolutionary Algorithms".

Main topics: mathematics, computer science, electronics, control, physics.

1997–2002

Diploma di Perito informatico (second grade secondary school with IT specialization)

ITIS "Luigi dell'Erba" (Istituto Tecnico Industriale Statale), Castellana Grotte (Italy)

Final Mark: 100/100

Main topics: mathematics, physics, chemistry, computer science, electronics

Skills

Programming languages / frameworks

Python (including Numpy, Scikit-learn, Pytorch, Tensorflow)
C / C++
Julia (including Knet for automatic differentiation)
Prolog (Yap-Prolog)
ROS (Robot Operating System)

Awards

PhD scholarship

IWT (government agency for Innovation by Science and Technology) scholarship for strategic base research (doctoraatsbeurs van strategisch basisonderzoek), 2012-2015

Best Paper awards

Best Student Paper - Machine Learning Journal Award:

D. Nitti, V. Belle, L. De Raedt. Planning in discrete and continuous Markov decision processes by probabilistic programming. European Conference on Machine Learning and Knowledge Discovery in Databases (ECML/PKDD 2015).

Personal (unpublished) projects

Differentiable assembly machine (python, Tensorflow): a simplified differentiable CPU model that extends “Neural Turing Machines”.
Deep reinforcement learning agents (Q-learning, policy gradient and other methods) applied to Gym-OpenAI environments.
Adaptive importance sampling.
Deep learning (Pytorch, Tensorflow): auto-encoders, CNN, RNN, meta-learning with RNN.

Some projects are publicly available at https://github.com/davidenitti/ML