MoroccoAI

Join us this Thursday, the 31st, for an insightful webinar on Atlas-Chat: Adapting Large Language Models for Low-Resource Moroccan Arabic Dialects, presented by Dr. Guokan Shang, Lead Research Scientist at MBZUAI France Lab, and Yousef Khoubrane, Graduate Student at UM6P. In this webinar, Guokan and Yousef will introduce Atlas-Chat, a collection of large language models specifically developed for Moroccan Arabic (Darija). They will discuss the construction of the dataset and present the evaluation of these models on several Darija benchmarks. 🗓 Save the date: Date: Thursday, the 31st Time: 7:00 PM Morocco time (GMT+1) 📝 The Zoom registration link is available in the comments below. We look forward to seeing you there! #MoroccoAI #Darija #LLM #Webinar

After successful editions featuring renowned speakers, including Nobel Prize laureate Geoffrey Hinton and Turing award winner Yoshua Bengio, we are delighted to announce the 4th edition of the MoroccoAI Annual Conference, which will take place virtually on December 4-5, 2024. The theme of this year is "Driving the Future of Innovation Through AI", where the goal goes beyond showcasing the latest AI advancements to discussing the impact AI will have on the innovation process itself. Contact us at conference@morocco.ai to get involved or if you would like to offer any suggestions. Stay tuned for more updates and registration details. For more information: www.morocco.ai #MoroccoAI #MoroccoAI2024 #AI #Innovation

#Neurips 2024 results are out and we congratulate all the authors (especially Moroccans who continue to show their excellence in AI research). The github repo https://lnkd.in/eEXJ95Zx is being updated with 2024 papers. Please do not hesitate to help us by adding/correcting the entries and sharing with others. This is also a reminder for our MoroccoAI Neurips prize for Morocco-affiliated authors ! please read the rules and fill the form (at the end of the readme file) to apply if you are affiliated to a Moroccan research/industry institute and have a paper accepted in Neurips main track this year !

We had the pleasure of hosting an insightful webinar on “Cancer Diagnosis with Quantum Machine Learning” delivered by Walid El Maouaki a Ph.D. student at Hassan II University in Casablanca, Morocco. During this webinar, Walid started by providing fundamental concepts of quantum computing and highlighted the key differences between classical machine learning and quantum machine learning. He explained the advantages quantum machine learning could bring to enhance decision-making in cancer diagnosis and improve healthcare services. Walid also demonstrated a practical application of quantum machine learning for cancer detection by exploring the use of Quantum SVM (QSVM) in prostate cancer detection. He introduced the key concepts behind QSVM and compared its performance to classical SVM models. 🎥 If you missed the webinar or would like to rewatch it, check the link to the recording in the comment section below. A big thanks to Walid for sharing his expertise and to all attendees #MoroccoAI #QuantumComputing #Webinar #QuantumMachineLearning

On the occasion of the Deep Learning Indaba 2024 that was held in Dakar this month, we met Walid El Maouaki who represented IndabaX Morocco in this event as he was the winner of the best IndabaXMorocco 2024 poster presentation. We are pleased to invite him for this month’s MoroccoAI webinar. 🚀 Join us this Friday the 27th, for an exciting webinar on Quantum Machine Learning for the enhancement of cancer diagnosis with our speaker Walid El Maouaki, a Ph.D. student at Hassan II University in Casablanca, Morocco, specializing in quantum machine learning applications and security. During this session, Walid will explore how Quantum Support Vector Machines (QSVM) can significantly improve the prediction accuracy of cancer detection, contributing to personalized medicine in oncology by accelerating detection processes and improving diagnostic precision. The webinar will also cover practical applications, challenges, and the future of integrating QML models into healthcare—highlighting their potential to transform patient care. 📅 Save the date: Date: Friday, the 27th Time: 8:00 PM Morocco time (GMT+1) 📝 Register here https://lnkd.in/ebtMHvMZ Looking forward to seeing you there! #MoroccoAI #QuantumMachineLearning #QML #CancerDiagnosis

On the occasion of the Deep Learning Indaba 2024 that was held in Dakar this month, we met Walid El Maouaki who represented IndabaX Morocco in this event as he was the winner of the best IndabaXMorocco 2024 poster presentation. We are pleased to invite him for this month’s MoroccoAI webinar. 🚀 Join us this Friday the 27th, for an exciting webinar on Quantum Machine Learning for the enhancement of cancer diagnosis with our speaker Walid El Maouaki, a Ph.D. student at Hassan II University in Casablanca, Morocco, specializing in quantum machine learning applications and security. During this session, Walid will explore how Quantum Support Vector Machines (QSVM) can significantly improve the prediction accuracy of cancer detection, contributing to personalized medicine in oncology by accelerating detection processes and improving diagnostic precision. The webinar will also cover practical applications, challenges, and the future of integrating QML models into healthcare—highlighting their potential to transform patient care. 📅 Save the date: Date: Friday, the 27th Time: 8:00 PM Morocco time (GMT+1) 📝 Register here https://lnkd.in/ebtMHvMZ Looking forward to seeing you there! #MoroccoAI #QuantumMachineLearning #QML #CancerDiagnosis

Last Wednesday, we had the pleasure of hosting an insightful webinar on “Algorithms with Untrusted Predictions” delivered by Ziyad Benomar, PhD student at ENSAE, France. During the webinar, Ziyad discussed online algorithms paradigm for decision-making under uncertainty, where decisions depend on unknown variables. He also presented a class of online algorithms called learning-augmented algorithms, which rely on machine learning predictions to design the algorithms, even if the predictions are inaccurate, highlighting  three critical criteria for these algorithms: consistency, robustness, and smoothness. Ziyad also presented his recent paper accepted at ICML 2024, where he introduced new efficient learning-augmented algorithms, which work with both partial and full predictions of unknown variables. His work is the first to yield a consistency-smoothness trade-off. Ziyad will be in Vienna to present his work at ICML2024 do not hesitate to reach out to him if you are also attending ! 🎥 If you missed the webinar or would like to watch it again, please check the link to the recording in the comment section below. A warm thanks to Ziyad for generously sharing his work and expertise, and to all attendees for making the webinar a success. #MoroccoAI #DecisionMaking #LearningAugmentedAlgorithms #ICML

Join us tomorrow for an engaging webinar on Algorithms with Untrusted Predictions with Ziyad Benomar, a PhD student at ENSAE, France, as he will be sharing his insights and findings on decision-making under uncertainty. In this era of machine learning, we often have predictions for unknown parameters in decision-making problems. But what if these predictions are inaccurate or even adversarial? During this webinar, Ziyad will introduce the "learning-augmented algorithms" paradigm and present his findings of his latest paper accepted at #ICML on non-clairvoyant scheduling to address this challenge. 📅 Save the date: Date: Wednesday, July 10th Time: 8PM (GMT+1), 9 PM (CET). 🔗Link to the paper: https://lnkd.in/esqxy8d4 Please take a minute to RSVP to receive event Zoom link and mark your calendar, 📝 link to register: https://lnkd.in/ei4b2bAx We're excited to have you join us for this illuminating session!

We had the pleasure of hosting an insightful webinar on “Multimodal AI for Radiology Applications” delivered by Kenza Bouzid, an Applied Researcher at Microsoft Health Futures in the Biomedical Imaging team, based in Cambridge, UK. During this webinar, Kenza first discussed the significant importance of grounded reporting by the example of a description of all findings in chest X-ray images, that can help radiologists in analyzing chest X-ray images, and how this improves communication with patients and enhances the comprehensiveness of reports. Kenza also introduced MAIRA-2, a CXR-specialized multimodal model for generating both grounded and non-grounded reports, detailing its architecture and comparing it with MAIRA-1. Additionally, she presented a novel evaluation framework that leverages the logical inference capabilities of LLMs for report generation, applicable to both text-only and text-and-grounding levels. 🎥 If you missed the webinar or would like to watch it again, please check the link to the recording in the comment section below. A warm thanks to Kenza for generously sharing her work and expertise, and to all attendees for making the webinar a success. #Morocco #MoroccoAI #Webinar #AI #Radiology #MultimodalAI

📢 Mark your calendars for June 19th! Join us for a compelling webinar on Multimodal AI for Radiology Applications with Kenza Bouzid, an Applied Researcher at Microsoft Health Futures in the Biomedical Imaging team, based in Cambridge, UK. Radiology reporting is a complex task requiring detailed image understanding, integration of multiple inputs, including comparisons with prior imaging, and precise language generation. This makes it an ideal area for the development and use of generative multimodal models. 📚 During this webinar, Kenza will detail MAIRA-2, a large multimodal model that combines a radiology-specific image encoder with a large language model (LLM). It is trained for the new task of grounded report generation on chest X-rays. She will also introduce RadFact, a novel evaluation framework leveraging the reasoning capabilities of LLMs to assess the factuality of individual generated sentences and the correctness of generated spatial localizations when present. 📅 Save the date: Date: Wednesday, June 19th Time: 8 PM (GMT+1), 9 PM (CET) 📝 Please take a minute to RSVP to receive the Zoom link and mark your calendar: https://lnkd.in/eAExnKwe Link to the paper: https://lnkd.in/eJhHMJbi We look forward to having you join us for this insightful session!