Sign In

Attilio Sbrana, PhD, MSc, MBA

Vice President @ Goldman Sachs | AI Research Scientist

Attilio Sbrana is a highly qualified professional currently serving as a Vice President in Artificial Intelligence and Quantitative Strategies at Goldman Sachs.2 His educational background is impressive, holding a PhD, MSc, and MBA, with his education including time at the Instituto Tecnológico de Aeronáutica (ITA).1

Career at Goldman Sachs

Sbrana joined Goldman Sachs in a leadership role, specifically as a Vice President focusing on Artificial Intelligence within the Quantitative Strategies team.2 He began this position on May 29, 2023, as indicated by his LinkedIn post announcing his new role.2

Professional Background

Prior to his current position, Sbrana's LinkedIn profile suggests he has experience as an AI Research Scientist.1 This background likely contributes to his expertise in applying artificial intelligence techniques to quantitative strategies in the financial sector.

LinkedIn Presence

Attilio Sbrana maintains an active professional presence on LinkedIn. His profile can be found under the username "attiliosbrana", which aligns with the information provided in the query.12

Sbrana's career trajectory demonstrates a strong blend of academic achievements and practical experience in the fields of artificial intelligence and quantitative finance, positioning him as a notable figure in the intersection of these domains at a leading global investment bank.

Highlights

Jan 3 · linkedin

Back at 200 West!

I’m happy to share that I’m starting a new position as Vice President, Artificial Intelligence, Quant Strats at Goldman Sachs!

Sep 22 · linkedin

🎉 Exciting Announcement: N-BEATS Perceiver Paper Published! 🎉

We're thrilled to share that our research paper, "N-BEATS Perceiver: A Novel Approach for Robust Cryptocurrency Portfolio Forecasting", has been published in the Springer Nature journal, Computational Economics!

🔍 About the Paper: Our study delves deep into a new groundbreaking method focused on univariate point-forecasting and portfolio selection. Drawing from an expansive dataset of over 4 million portfolio test samples, which translates to 192 million point forecasts, the results are evident: our Transformer-based N-BEATS Perceiver model surpasses both conventional and modern-day forecasting techniques. When we shift our lens to portfolio selection, the model's risk resilience stands out, especially when evaluated in an MCDM setting using the TOPSIS method.

🏋️ On Hardcore Experimentation: Integral to our research philosophy is the firm belief that robust benchmarks demand harder and larger test sets. It is our earnest endeavor to elevate the standard of benchmarks and guide the community towards authentic and effective forecasting methodologies.

💡 A Glimpse into Our Research: On page 27, we've juxtaposed results of 40 univariate forecasting methods for various cryptocurrency portfolios (ie N-HiTS, N-BEATS, ARIMA, ETS, theta, etc). Our research underscores the pivotal role of Transformer methods and RNNs as, potentially, the only effective mechanisms that offer forecasts superior to naive methods in the case of cryptocurrencies. All our findings are reproducible with open-source data for continued exploration.

🧠 Deep Learning's Mysterious Cross-Learning: The overarching essence of our research is that of training a deep learning model to cross-learn, self-supervised, many portfolios over many years of training data. This resonates with the theory of deep learning as a Kolgomorov compressor with finite compute and memory. The model improves through capturing algorithmic mutual information, mathematically expressed as: I_K(series_X; series_Y) = K(series_X) - K(series_X|series_Y)

This is likely the reason why we will continue to see deep learning shining as the best forecasting method going forward, as there is little regret to K(series_X|series_Y).

For those who wish to dive deeper, you can find our paper here: https://lnkd.in/dmdpkrfb

Regrettably, it's not open access, but I have a link with limited free copies and I'm always open to discussing our findings and exchanging thoughts!

Thank you for your continued support, and we look forward to sharing more such milestones with our community. 💼📚

Related Questions

What are Attilio Sbrana's main responsibilities at Goldman Sachs?
What specific projects has Attilio Sbrana worked on in AI at Goldman Sachs?
How did Attilio Sbrana transition from ITA to a career in AI at Goldman Sachs?
What are some notable achievements of Attilio Sbrana in the field of artificial intelligence?
What is Attilio Sbrana's educational background and how has it influenced his career?
Attilio Sbrana, PhD, MSc, MBA
Add to my network

Location

New York, New York, United States