Evolving the Topology and the Weights of Neural Networks Using a Dual Representation

Evolutionary computation is a class of global search techniques based on the learning process of a population of potential solutions to a given problem, that has been successfully applied to a variety of problems. In this paper a new approach to the construction of neural networks based on evolutionary computation is presented. A linear chromosome combined to a graph representation of the network are used by genetic operators, which allow the evolution of the architecture and the weights simultaneously without the need of local weight optimization. This paper describes the approach, the operators and reports results of the application of this technique to several binary classification problems.     

A decision-support system for the design and management of warehousing systems

The issue of material handling involves the design and operative control of warehousing systems (i.e., distribution centres), which allow matching vendors and demands, smoothing with seasonality, consolidating products and arranging distribution activities. Warehousing systems play a crucial role in providing efficiency and customer satisfaction. The warehouse design entails a wide set of decisions, which involve layout constraints and operative issues that seriously affect the performances and the overall logistics costs.     

Biohybrid Bovine Bone Matrix for Controlled Release of Mesenchymal Stem/Stromal Cell Lyosecretome: A Device for Bone Regeneration

SmartBone^® (SB) is a biohybrid bone substitute advantageously proposed as a class III medical device for bone regeneration in reconstructive surgeries (oral, maxillofacial, orthopedic, and oncology). In the present study, a new strategy to improve SB osteoinductivity was developed. SB scaffolds were loaded with lyosecretome, a freeze-dried formulation of mesenchymal stem cell (MSC)-secretome, containing proteins and extracellular vesicles (EVs). Lyosecretome-loaded SB scaffolds (SBlyo) were prepared using an absorption method. A burst release of proteins and EVs (38% and 50% after 30 min, respectively) was observed, and then proteins were released more slowly with respect to EVs, most likely because they more strongly adsorbed onto the SB surface. In vitro tests were conducted using adipose tissue-derived stromal vascular fraction (SVF) plated on SB or SBlyo. After 14 days, significant cell proliferation improvement was observed on SBlyo with respect to SB, where cells filled the cavities between the native trabeculae. On SB, on the other hand, the process was still present, but tissue formation was less organized at 60 days. On both scaffolds, cells differentiated into osteoblasts and were able to mineralize after 60 days. Nonetheless, SBlyo showed a higher expression of osteoblast markers and a higher quantity of newly formed trabeculae than SB alone. The quantification analysis of the newly formed mineralized tissue and the immunohistochemical studies demonstrated that SBlyo induces bone formation more effectively. This osteoinductive effect is likely due to the osteogenic factors present in the lyosecretome, such as fibronectin, alpha-2-macroglobulin, apolipoprotein A, and TGF-β.     

Nano-to-macroporous TiO2 (anatase) by cold sintering process

Cold Sintering Process (CSP) was applied on commercial nanopowders to produce nanostructured TiO₂ anatase with nano-to-macro porosity. Nanoporous TiO₂ based materials were obtained by applying CSP at 150 °C and pressures up to 500 MPa on three TiO₂ nanopowders with different specific surface area (s.s.a. = 50, 90 and 370 m²/g), using water as transient aqueous environment. Although TiO₂ is insoluble in water, a density of 68% and s.s.a. = 117 m²/g were achieved from the powder with the highest specific surface area. A post annealing process at 500 °C increased the density up to 73% with a s.s.a. = 59 m²/g, and the crystallites dimensions passed from 110 Å in the powder to 130 Å in CSP material and 172 Å after post annealing. Finally, macroporosity was produced by using thermoplastic polymer beads as sacrificial templates within TiO₂ nanopowder during CSP, followed by a debonding at 500 °C.     

Linear optimal control strategies for production systems with a discrete-event demand pattern

A class of production systems are considered in this paper with the aim of determining closed-loop strategies providing the production effort as functions of the system state. The key feature of the considered class of systems is that the demand and the way it is satisfied are asynchronous sequences of part requests and instantaneous withdrawals, respectively, generated by discrete-event processes, whereas the production process has continuous-time dynamics. The optimization problem, whose objective is to minimize inventory and lateness costs, is restated as an optimal control problem by exploiting some structural properties of the optimal solution. Then, optimal closed-loop strategies are found with dynamic programming techniques.     

Shot scale distribution in art films

The scale of shot, i.e. the apparent distance of the camera from the main subject of a scene, is one of the main stylistic and narrative functions of audiovisual products, conveying meaning and inducing the viewer’s emotional state. The statistical distribution of different shot scales in a film may be an important identifier of an individual film, an individual author, and of various narrative and affective functions of a film. In order to understand at which level shot scale distribution (SSD) of a movie might become its fingerprint, it is necessary to produce automatic recognition of shot scale on a large movie corpus. In our work we propose an automatic framework for estimating the SSD of a movie by using inherent characteristics of shots containing information about camera distance, without the need to recover the 3D structure of the scene. In the experimental investigation, the comparison of obtained results with manual SSD annotations proves the validity of the framework. Experiments conducted on movies by Michelangelo Antonioni taken from different stylistic periods (1950–57, 1960–64, 1966–75, 1980–82) show a strong similarity in shot scale distributions within each period, thus opening interesting research lines regarding the possible aesthetic and cognitive sources of such a regularity.     

Caffeic Acid Enhances the Anti-Leukemic Effect of Imatinib on Chronic Myeloid Leukemia Cells and Triggers Apoptosis in Cells Sensitive and Resistant to Imatinib

Among the phenolic acids tested on the K562 cell line, a model of chronic myeloid leukemia (CML), caffeic acid (CA) was biologically active on sensitive and imatinib (IM)-resistant cells at micro-molar concentration, either in terms of reduction of cell proliferation or triggering of apoptosis. The CA treatment provoked mitochondrial membrane depolarization, genomic DNA fragmentation and phosphatidylserine exposure, hallmarks of apoptosis. Cell cycle analysis following the treatment with comparable cytotoxic concentrations of IM or CA showed marked differences in the distribution profiles. The reduction of cell proliferation by CA administration was associated with increased expression of two cell cycle repressor genes, CDKN1A and CHES1, while IM at a cytotoxic concentration increased the CHES1 but not the CDKN1A expression. In addition, CA treatment affected the proliferation and triggered the apoptosis in IM-resistant cells. Taken together, these data suggested that CA induced the anti-proliferative effect and triggered apoptosis of CML cells by a different mechanism than IM. Finally, the combined administration of IM and CA at suboptimal concentrations evidenced a synergy of action in determining the anti-proliferative effect and triggering apoptosis. The ability of CA to potentiate the anti-leukemic effect of IM highlighted the nutraceutical potential of CA in CML.     

Elastic modulus of nanocrystalline cemented carbide

The purposes of this work were to obtain the accurate elastic modulus of the nanocrystalline WC–Co cemented carbides, and to propose the mechanism for the difference of elastic modulus between the nanocrystalline and conventional polycrystalline cemented carbides. The nanocrystalline cemented carbide was prepared by spark plasma sintering (SPS) technique. The conventional polycrystalline cemented carbides were prepared by SPS and sinter-HIP techniques as references, respectively. The sintered cemented carbides were characterized by X-ray diffractometry, scanning electron microscopy and the transmission electron microscopy with precession electron diffraction technology. The elastic modulus was obtained by averaging the values measured with the continuous stiffness measurement method of the nanoindentation technology. The results show that the nanocrystalline cemented carbide has a relatively low modulus, which could be attributed to the more interface area and higher fraction ratio of the hcp cobalt phase caused by the rapid heating and cooling process during SPS.