Highlight from the ADIA Lab Casual Discovery Challenge in Abu Dhabi, where Andre Franca, PhD was co-awarded second place! 🎉🎊
Check out Andre briefly walk through their submission and its unique approach. An in-depth look at the methodology behind the winning project and more insights into #CausalDiscovery coming soon!! 👀
Hi I am Andre Franca CTO at Ergodic.ai and I'm here to talk about the solution that Alexis Gassman and I did for our submission. So we we did a mix of two different approaches. The first approach was using some or modifications on already established causal discovery algorithms. I'll go into more detail. And the second approach was using deep learning method and novel deep learning method. And then at the end we built our own infrastructure to actually assemble these together. So to go into more detail, the right off the bat, most of the out-of-the-box solutions for causal discovery just didn't work very well on this data because this data is full of linearities, nonlinearities. Of course, the the noise of the data was kind of obscure because we're only looking at the copula. And then anyway, there were a lot of technical challenges that made that out-of-the-box calls the discovery methods not work really well. So our first part of the solution. They come up with our own pipeline that use PC algorithm using a nonlinear. Linear transformations on the data to actually create a conditional independence test that was coming up with a much better skeleton that were traditional PC was was was doing so then we we use our also our own modification of an Astar search algorithm again with a completely different scoring function to actually come up with orientations on top of that skeleton. But that was just enough to get to 68 percent or something, which is. Still a bit far from from, you know, where the state-of-the-art algorithms in this competition ended up. So then we, so Alexis and I, then we use the deep learning approach, which is based on the model called Avicii that uses a variational auto encoder to fine tune on the data to actually come up with estimates of the calls that we have. What was interesting is that there was a lot of fatality between these two approaches, which means that there was some opportunity to come up with an assembly method. And if you think about it, actually you don't need to restrict yourself to these two approaches. You can throw in basically any causal discovery algorithm that you want on top of an assembler. And then this ensemble of what you really do is that it will look at the characteristics of the data. It will look at how many features you have, what's the covariance properties of the covariance matrix, some random matrix theory properties of the data to come up with a final. Probability distribution and each one of these edges. And then what you can do afterwards is to reconstruct the graph on top of these assembler and come up with a with a final estimate. And this is what we did. So ultimately, right, it was really, really challenging because it's a very nascent field. There aren't a lot of out-of-the-box solutions available. So we actually had to go and start from scratch and think really carefully about what's present in the data and how we can make that surface out in the algorithms that we were building. So thank you very much. Everyone.
On my way to IPMU (https://lnkd.in/esY_aWKd) in gorgeous Lisboa, a conference with a friendly atmosphere to exchange about uncertainty. In order of appearance, we will talk about:
* Using robust kernels to process signal and more particularly to remove noisy pixels from images, using imprecise probabilistic models. A work done with Frederic Comby and Olivier Strauss from the LIRMM
* Decombining belief functions efficiently, an essential tool if we want to correct potential inconsistencies between pieces information. This work was mainly done by Daira Pinto Prieto together with Ronald de Haan, and is a follow-up of Daira visit funded by the SAFE AI chair.
* Propagating uncertainties in geospatial problems, with a specific focus on predicting pollutants in soils. This work is part of Priscillia Labourg thesis, and was done in collaboration with romain guillaume, Benjamin Quost, Jérémy Rohmer and Stephane Belbeze. Priscillia thesis is funded by the project HOUSES which is led by the BRGM institute.
passing on an opportunity
"We are reaching out to ask for your organization’s help to spread the word that the Airport Cooperative Research Program (ACRP) is now accepting Problem Statements. Each year ACRP relies on airport stakeholders to identify key industry issues to guide the focus of its forthcoming research projects.
From now until April 8th, stakeholders can submit problem statements for new research projects to take place in the FY2025 program year. ACRP is an industry-driven, applied research program that develops near-term, practical solutions to airport challenges. Problem statements inspire research that leads to the publication of no-cost guidelines, tools, best practices, databases, and/or other resources that will support practitioners in addressing current industry challenges.
A call for proposals worth €83.7 million in EU funding specifically dedicated to accelerate the green transition in the Mediterranean through collaborative solutions: this is the key outcome of the recent meeting of the Monitoring Committee, Interreg NEXT MED’s decision-making body, held on July 17-18, 2024, in Athens.
https://lnkd.in/eZU9Q4QM
Attending the conference "Italia Chiama Europa, il futuro degli studentati al 2030" organized by Scenari Immobiliari at Fondazione Giangiacomo Feltrinelli in Milan with Massimo Roj, Gabriele Cerminara and Francesco Gallina on behalf of Progetto CMR International, which has now numerous student residence projects under its portfolio.
A conference in collaboration with Re.Uni (Residenze Universitarie Italiane) that introduced the Report "Lo student housing da mercato di nicchia a comparto maturo" unveiling data and banchmark on the market of Residences for University Students.
It was also an important opportunity to address the next steps of the PNRR funds by analyzing the resources allocated to university housing with an analysis of the private investor market.
#event#studenthousing#scenariimmobiliari#reports#progettocmr#progettocmrintenational
"Delighted to share my certificate of participation in the Indian Knowledge System Project Expo! This expo was an incredible opportunity to learn from and showcase innovative cultural projects. I gained valuable insights and experiences that will help me grow creativity and knowledge.
#IndianKnowledgeSystem#ProjectExpo#LearningAndDevelopment"
🔔👥Did you miss our Final Conference? Maybe after this recap you’ll want to check out the complete recording!
For the 60 attendees (40 that joined the conference in presence and 20 online) it was the perfect opportunity to explore the project’s accomplishments, contributions and impact on the tourism sector.
They had the chance to meet and get to know each other, learn the point of view and perspectives of other entities in the field, acquire hands-on expertise in enhancing digital tools for tourism SMEs, discover exemplary strategies in digital tourism and develop professional networks with entities from the tourism industry.
The conference agenda focused on four main sessions:
The welcoming remarks, that provided the participants with an overview of the project scope, the conference program and the European context within which the project developed. The session saw on the stage Ignasi Papell Garcia, Eurecat Territorial Manager for Southern Catalonia, Eduard Rovira, First Vice-President of Costa Daurada and Terres de l'Ebre Tourist Board, Ignacio de las Cuevas, RESETTING project’s manager and Manager of Centre of Excellence in Tourism Innovation at Eurecat - Technology Centre, and finally Ricardo Aguado, RESETTING project officer and adviser from European Innovation Council and SMEs Executive Agency (EISMEA).
A roundtable and Knowledge Exchange Session, that was directed by Eva Katsaraki from the Heraklion Development Agency and involved five of the SMEs that participated in the project’s open calls: Rentalmar, Seven Rooms and Flat, Aftersails-Nefalo, BlueGeo and Albania my Way, that shared their experiences with the RESETTING project and the direct impacts it had on their businesses.
Two practical and inspiring speeches, the first involving Àurea Rodríguez, Interim director at EIT Culture & Creativity, that delved into the topic of decolonizing value chains, and the second involving Gian Franco Mercado, hospitality specialist, that showed the audience the tools ans skills needed to unlcok the power of AI for tourism companies.
A roundtable with the other COSME projects dedicated to the tourism sector, that brought together representatives from the other sister projects:
Sergio Moreno Gil from Euroemotur, Yeimy Ospina Valencia from Tourism 4.0, Caitlín Breathnach from ST3ER and Ricardo Aguado from EISMEA.
During the day, the attendees could also experience a series of live demo sessions of the Smart Tourism Tools developed by the project’s teams led by Sandra Loureiro, Full Professor at BRU- ISCTE and Fernando Brito e Abreu, Associate Professor at DCTI - ISCTE, that were supported by professors Daniela Langaro, Rui Marinheiro and by the rest of their valiant teams.
If you missed the event you can find the full conference here 👉https://lnkd.in/dii9XBxp
Or you can wait for the recap video that will be published soon!
Southern Catalonia ICT ClusterAlbanian TripFEHT
Integrating eSIM as an ancillary service can be daunting without telecom expertise, technical know-how, or a substantial budget. However, embracing eSIM isn’t merely an upgrade; it’s a strategic move every hotel should consider.
Discover how Vista Grand overcame these challenges and achieved success in just three weeks by diving into our full case study!
Meet us at the GBTA Convention next week to discover how you can elevate your business with Zetexa Global!
#Hospitality#Zetexa#eSIM#Innovation#AncillaryRevenue#CaseStudy#ZetSIM
Thanks for pushing research on the air cargo topic!! You may want to consider the Special Issue in the Journal of Shipping and Trade as a potential outlet for your papers: https://lnkd.in/g-wdVjkF
Aix-Marseille-Provence Metropolis has just been ranked in the Top 10 of European Cities and Regions of the Future (Top 10 Mid-Sized European Regions of the Futur 2024 - FDI Strategy (Foreign Direct Investment Strategy) by fDi Intelligence/Financial Times
https://lnkd.in/e82JNeZShttps://lnkd.in/eqf6W68c
This Ranking highlights the work made by the Metropolitan Authority and its partners in terms of :
=> Competitiveness, to design and build different & innovative offers to welcome investments of companies in some strong fields of our Territory (Aeronautics & Mecanics; Environment & Energy; Health & Biotech; Maritime/Naval & Logistics; Digital, Creative Industries & Microelectronics; Tourism)
=> Attractiveness, with our shared Brand, One Provence, to gather together the local actors towards a common ambition to the World : Aix-Marseille-Provence is the place where we are building the City of the Future, which consists in a sustainable, productive and qualitative port Metropolis in the heart of the Mediterranean
All these Assets, we will present them on the MIPIM Exhibition in Cannes from 12th to 15th March (Stand R7-E57, Hall Riviera) with our partners :
https://lnkd.in/e4UvQ-pt
#AivpWorldConference Discover the programme 👉 https://swll.to/zq3kV
"𝐎𝐩𝐞𝐧 𝐏𝐢𝐞𝐫𝐬: 𝐒𝐭𝐞𝐞𝐫𝐢𝐧𝐠 𝐟𝐥𝐨𝐰𝐬 𝐛𝐞𝐭𝐰𝐞𝐞𝐧 𝐩𝐞𝐨𝐩𝐥𝐞, 𝐩𝐥𝐚𝐧𝐞𝐭, 𝐚𝐧𝐝 𝐩𝐨𝐫𝐭 𝐜𝐢𝐭𝐢𝐞𝐬"
🗓 27-29 November 2024 📍 Lisbon (Portugal)
Piers are the ultimate interface between land and water. They are also the connection between port cities and the planet. Historically, their accessibility was a metaphor for the evolution of the port-city relationship. They changed from being accessible in the past, to gradually become off limits. Fortunately, there is a tendency to give back this access and reopen the piers for the people, inaugurating a new kind of port-city relationship. . But how permeable are these relationships today? What flows are shaping these interactions? How are port cities presenting local answers to global trends? What bridges? And how are ports reopening their piers?
Porto de Lisboa
Well done! Like the name Ergodic - stable, repeatable signals/ strategies