Zobrazeno 1 - 10
of 61
pro vyhledávání: '"Rozza, Alessandro"'
Autor:
Stradi, Francesco Emanuele, Cipriani, Filippo, Ciampiconi, Lorenzo, Leonardi, Marco, Rozza, Alessandro, Gatti, Nicola
We address the challenging problem of dynamically pricing complementary items that are sequentially displayed to customers. An illustrative example is the online sale of flight tickets, where customers navigate through multiple web pages. Initially,
Externí odkaz:
http://arxiv.org/abs/2407.05793
Most state-of-the-art machine learning techniques revolve around the optimisation of loss functions. Defining appropriate loss functions is therefore critical to successfully solving problems in this field. We present a survey of the most commonly us
Externí odkaz:
http://arxiv.org/abs/2301.05579
Contextual bandits can solve a huge range of real-world problems. However, current popular algorithms to solve them either rely on linear models, or unreliable uncertainty estimation in non-linear models, which are required to deal with the explorati
Externí odkaz:
http://arxiv.org/abs/2210.06302
Publikováno v:
IEEE Transactions on Image Processing, 31 (2022) 5009-5024
The aesthetic quality of an image is defined as the measure or appreciation of the beauty of an image. Aesthetics is inherently a subjective property but there are certain factors that influence it such as, the semantic content of the image, the attr
Externí odkaz:
http://arxiv.org/abs/2111.04647
With the rise in use of social media to promote branded products, the demand for effective influencer marketing has increased. Brands are looking for improved ways to identify valuable influencers among a vast catalogue; this is even more challenging
Externí odkaz:
http://arxiv.org/abs/2107.13943
Autor:
Manessi, Franco, Rozza, Alessandro
Self-supervised learning is currently gaining a lot of attention, as it allows neural networks to learn robust representations from large quantities of unlabeled data. Additionally, multi-task learning can further improve representation learning by t
Externí odkaz:
http://arxiv.org/abs/2011.07267
Autor:
Manessi, Franco, Rozza, Alessandro
Publikováno v:
2018 24th International Conference on Pattern Recognition (ICPR), Beijing, 2018, pp. 61-66
In the last decade, an active area of research has been devoted to design novel activation functions that are able to help deep neural networks to converge, obtaining better performance. The training procedure of these architectures usually involves
Externí odkaz:
http://arxiv.org/abs/1801.09403
Publikováno v:
2018 24th International Conference on Pattern Recognition (ICPR), Beijing, 2018, pp. 657-664
In this work we present a method to improve the pruning step of the current state-of-the-art methodology to compress neural networks. The novelty of the proposed pruning technique is in its differentiability, which allows pruning to be performed duri
Externí odkaz:
http://arxiv.org/abs/1712.01721
Publikováno v:
The final version has been published in Elsevier Pattern Recognition, August 2019
Many different classification tasks need to manage structured data, which are usually modeled as graphs. Moreover, these graphs can be dynamic, meaning that the vertices/edges of each graph may change during time. Our goal is to jointly exploit struc
Externí odkaz:
http://arxiv.org/abs/1704.06199
We present a theoretically grounded approach to train deep neural networks, including recurrent networks, subject to class-dependent label noise. We propose two procedures for loss correction that are agnostic to both application domain and network a
Externí odkaz:
http://arxiv.org/abs/1609.03683