This paper presents a comparative analysis of deep learning strategies for detecting hypertensive retinopathy from fundus images, a central task in the HRDC challenge~\cite{qian2025hrdc}. We investigate three distinct approaches: a custom CNN, a suite of pre-trained transformer-based models, and an AutoML solution. Our findings reveal a stark, architecture-dependent response to data augmentation. Augmentation significantly boosts the performance of pure Vision Transformers (ViTs), which we hypothesize is due to their weaker inductive biases, forcing them to learn robust spatial and structural features. Conversely, the same augmentation strategy degrades the performance of hybrid ViT-CNN models, whose stronger, pre-existing biases from the CNN component may be "confused" by the transformations. We show that smaller patch sizes (ViT-B/8) excel on augmented data, enhancing fine-grained detail capture. Furthermore, we demonstrate that a powerful self-supervised model like DINOv2 fails on the original, limited dataset but is "rescued" by augmentation, highlighting the critical need for data diversity to unlock its potential. Preliminary tests with a ViT-Large model show poor performance, underscoring the risk of using overly-capacitive models on specialized, smaller datasets. This work provides critical insights into the interplay between model architecture, data augmentation, and dataset size for medical image classification.

In this work we investigate different machine learning based strategies for denoising raw simulation data from ProtoDUNE experiment. Proto-DUNE detector is hosted by CERN and it aims to test and calibrate the technologies for DUNE, a forthcoming experiment in neutrino physics. Our models leverage deep learning algorithms to make the first step in the reconstruction workchain, which consists in converting digital detector signals into physical high level quantities. We benchmark this approach against traditional algorithms implemented by the DUNE collaboration. We test the capabilities of graph neural networks, while exploiting multi-GPU setups to accelerate training and inference processes.

AI (ML) has been utilized for a long time in different industries to drive new business, increase productivity, reduce risk and improve consumer satisfaction. However, within data management, widespread adoption still can't seem to progress. One issue is that use cases and capacities of ML related to data management are not constantly comprehended by operational teams. Another is that the undeniable use cases require high levels of accuracy, while the accuracy of ML methods is as of now observed as hard to anticipate. Above all, there is a strong everyday spotlight on delivering cleansed data to downstream applications, for example, risk, trade support, and compliance engines, leaving little time to improve or set out on apparent, large undertakings.

Nowadays, users open multiple accounts on social media platforms and e-commerce sites, expressing their personal preferences on different domains. However, users' behaviors change across domains, depending on the content that users interact with, such as movies, music, clothing and retail products. In this paper, we propose an adaptive deep learning strategy for cross-domain recommendation, referred to as ADC. We design a neural architecture and formulate a cross-domain loss function, to compute the non-linearity in user preferences across domains and transfer the knowledge of users' multiple behaviors, accordingly. In addition, we introduce an efficient algorithm for cross-domain loss balancing which directly tunes gradient magnitudes and adapts the learning rates based on the domains' complexities/scales when training the model via backpropagation. In doing so, ADC controls and adjusts the contribution of each domain when optimizing the model parameters. Our experiments on six publicly available cross-domain recommendation tasks demonstrate the effectiveness of the proposed ADC model over other state-of-the-art methods. Furthermore, we study the effect of the proposed adaptive deep learning strategy and show that ADC can well balance the impact of the domains with different complexities.

Floating Content (FC) is a communication paradigm for the local dissemination of contextualized information through D2D connectivity, in a way which minimizes the use of resources while achieving some specified performance target. Existing approaches to FC dimensioning are based on unrealistic system assumptions that make them, highly inaccurate and overly conservative when applied in realistic settings. In this paper, we present a first step towards the development of a cognitive approach to efficient dynamic management of FC. We propose a deep learning strategy for FC dimensioning, which exploits a Convolutional Neural Network(CNN) to efficiently modulate over time the resources employed by FC in a QoS-aware manner. Numerical evaluations show that our approach achieves a maximum rejection rate of3%, and resource savings of 37.5% with respect to the benchmark strategy

Six months ago, I wrote that "The West is Unaware of the Deep Learning Sputnik moment". It turns out mathematician Cedric Villani began a 6 month journey to learn all that needs to be learned about Deep Learning. He is the author of a report that France will use to drive their future Deep Learning strategy. Villani's report is titled "For a Meaningful Artificial Intelligence: Towards a French and European Strategy". Although Villani uses the term "Artificial Intelligence" to appeal to a much wider audience, he is actually responding to the "particular" developments of Deep Learning.

A public report by Harvard reveals how unprepared the US Military is when it comes to the Artificial Intelligence (AI) technology known as Deep Learning. The study by Harvard's Kennedy center was published in July 2017, written by Greg Allen Taniel Chan, and was conducted with funding from IARPA. The research is titled "Artificial Intelligence and National Security". I've written about the many tribes of AI and about the use of the term AI being too ambiguous and meaning too many things to too many people. Where do we find Deep Learning in this report from Harvard?

Deep Learning is a technology that is as disruptive as mobile computing or the world wide web that came before it. Yet, most enterprises have no strategy on what to do. This is perplexing given that Deep Learning most hyped up slogan is "The Last Invention of Man". It boils down to one simple fact, enterprises don't understand Deep Learning. To make it even worse, they can't possibly understand the current wave of Artificial Intelligence (AI) developments if they don't understand Deep Learning.