Our institute welcomes as interns students in computer science, mathematics, physics, engineering, or economics, providing opportunities such as:
- – Starting a career in machine learning, currently the hottest topic in IT;
- – Being part of an international, multidisciplinary research environment;
- – Learning to do scientific research;
– Studying the market for a machine-learning-based startup;
– Creating the background necessary to develop a bachelor or master thesis in collaboration with us;
Deep Computational Intelligence Group (Răzvan Florian)
Deep learning / machine learning
Examples of internship projects:
- – Automatically generating HTML code using deep recurrent neural networks;
– Categorizing web pages based on their style and structure;
– Semantic clustering of website elements;
– Code autocompletion (e.g., the CodRep competition);
– Training deep neural networks that automatically generate interior designs;
– Competing in the AI Driving Olympics;
– Training spiking neural networks for benchmark reinforcement learning tasks;
– Documenting the physics behind Thyrix, a 2D simulator designed for artificial intelligence research (suitable for physics / mathematics students).
Marketing for startups / entrepreneurship
Suitable for economics students (see online course about startups):
- – Study the product-market fit for automated generation of web pages;
- – Study the product-market fit for automated generation of interior designs.
Physics / mathematics
- – Documenting the physics behind Thyrix, a 2D simulator designed for artificial intelligence research (suitable for physics / mathematics students).
Machine Learning and Optimization Group (Luigi Malagò)
Examples of internship projects:
- – Automatic image generation based on deep learning: learn how to train neural networks able to automatically generate images using variational auto-encoders;
– Stochastic optimization using deep learning: find the minimum of high-dimensional non-convex functions using generative models based on deep neural networks;
– Deep learning for sound classification: learn how to train convolutional neural networks and recurrent neural networks to recognize and classify different types of sounds, such as music instruments and environment sounds;
– Deep neural networks for text analysis: learn how to train neural networks to solve word analogies through word embeddings, such as “queen is to king, as woman is to ?”
– Deep learning for physics: learn how deep neural networks can be useful to solve optimization and inference tasks in physics, such as the estimation of the parameters of cosmological models or finding the minimum of quantum many-body problems.
How to apply
To apply, please fill this form no later than June 15, 2018.