Polytechnic University of Milan

About: Polytechnic University of Milan is a education organization based out in Milan, Italy. It is known for research contribution in the topics: Computer science & Finite element method. The organization has 18231 authors who have published 58416 publications receiving 1229711 citations. The organization is also known as: PoliMi & L-NESS.

Bisphenol A, nonylphenols, benzophenones, and benzotriazoles in soils, groundwater, surface water, sediments, and food: a review

Abstract: Contaminants of emerging concern (CECs) are not commonly monitored in the environment, but they can enter the environment from a variety of sources. The most worrying consequence of their wide use and environmental diffusion is the increase in the possible exposure pathways for humans. Moreover, knowledge of their behavior in the environment, toxicity, and biological effects is limited or not available for most CECs. The aim of this work is to edit the state of the art on few selected CECs having the potential to enter the soil and aquatic systems and cause adverse effects in humans, wildlife, and the environment: bisphenol A (BPA), nonylphenol (NP), benzophenones (BPs), and benzotriazole (BT). Some reviews are already available on BPA and NP, reporting about their behavior in surface water and sediments, but scarce and scattered information is available about their presence in soil and groundwater. Only a few studies are available about BPs and BT in the environment, in particular in soil and groundwater. This work summarizes the information available in the literature about the incidence and behavior of these compounds in the different environmental matrices and food. In particular, the review focuses on the physical-chemical properties, the environmental fate, the major degradation byproducts, and the environmental evidence of the selected CECs.

Biomedical Applications of Shape Memory Alloys

Abstract: Shape memory alloys, and in particular NiTi alloys, are characterized by two unique behaviors, thermally or mechanically activated: the shape memory effect and pseudo-elastic effect. These behaviors, due to the peculiar crystallographic structure of the alloys, assure the recovery of the original shape even after large deformations and the maintenance of a constant applied force in correspondence of significant displacements. These properties, joined with good corrosion and bending resistance, biological and magnetic resonance compatibility, explain the large diffusion, in the last 20 years, of SMA in the production of biomedical devices, in particular for mini-invasive techniques. In this paper a detailed review of the main applications of NiTi alloys in dental, orthopedics, vascular, neurological, and surgical fields is presented. In particular for each device the main characteristics and the advantages of using SMA are discussed. Moreover, the paper underlines the opportunities and the room for new ideas able to enlarge the range of SMA applications. However, it is fundamental to remember that the complexity of the material and application requires a strict collaboration between clinicians, engineers, physicists and chemists for defining accurately the problem, finding the best solution in terms of device design and accordingly optimizing the NiTi alloy properties.

Parallel Recurrent Neural Network Architectures for Feature-rich Session-based Recommendations

Abstract: Real-life recommender systems often face the daunting task of providing recommendations based only on the clicks of a user session. Methods that rely on user profiles -- such as matrix factorization -- perform very poorly in this setting, thus item-to-item recommendations are used most of the time. However the items typically have rich feature representations such as pictures and text descriptions that can be used to model the sessions. Here we investigate how these features can be exploited in Recurrent Neural Network based session models using deep learning. We show that obvious approaches do not leverage these data sources. We thus introduce a number of parallel RNN (p-RNN) architectures to model sessions based on the clicks and the features (images and text) of the clicked items. We also propose alternative training strategies for p-RNNs that suit them better than standard training. We show that p-RNN architectures with proper training have significant performance improvements over feature-less session models while all session-based models outperform the item-to-item type baseline.