Leonardo Dipilato
Community Lead @ GDG Pisa; Design Engineer @ Cubit
Actions
Computational Fluid Dynamics engineer, Data Analyst, and BIG nerd.
Leonardo has been doing freelance data analysis for years, doing trend analysis for market prediction, sentiment analysis on event feedback, and explainable AI for high-stakes applications. In recent years, he's also been applying Machine Learning models to physics applications, such as modal decomposition, sparse dynamics, and physics-informed reduced-order modeling.
Leonardo's Pokédex entry says: if he's not solving supercavitating flows around hydrofoils, wondering whether there's any bias in a dataset, or doing random Leetcode/Project Euler questions, you'll find him organizing community tech events or attending them. Approach with some food and you'll be best friends!
Topics
No Free Lunch Theorem: Machine Learning 101
The No Free Lunch Theorem states that there exists no learning algorithm better than any other given one, thus making it necessary for a ML Engineer to know their ways around multiple ways to tackle a problem.
In this talk we'll have a look at common algorithms and models that a ML Engineer should know and when to actually use them.
Neural Networks for Humans: demystifying the math
Neural Networks are, more often than not, treated as black boxes: there’s this strange idea of them copying the human brain in such a complex way that no human could ever understand, but that couldn’t be farther from the truth. Neural Networks are actually, conceptually, quite simple in how they work and make up for their simplicity by adding more and more stuff on top of them.
Knowing how Neural Networks work is a must for anyone who wants to work in the field of Machine Learning, but even as a bystander having a clear mental image of how stuff works can open your mind to new concepts.
In this workshop we’ll completely demystify how Neural Networks of all kinds work, not by completely disregarding the math side of things, but by actually using it in a human way (instead of cold artificial matrix multiplications).
JavaScript Shenanigans
JavaScript is an extremely common language whose core was developed in 10 days in 1995. Doing things with such a haste left some troubles in the core language that haven't been fixed.
In this talk, we'll look at some snippets of code to highlight some of the unusualities that JavaScript brings.
How to build your own Chess, Go and Shogi champion
In 2017, Google’s AlphaZero managed to beat Stockfish, the world’s strongest chess player (incidentally, an AI), using deep neural networks and a general game playing algorithm that can also play Go and Shogi, two other extremely complex games, which is similar to a human being the champion of all three games at the same time. In this talk we’ll cover all the techniques behind AlphaZero in order to fully understand how it works.
GDG DevFest Pescara 2019 Sessionize Event
DevFest Levante 2019 Sessionize Event
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top