-
evoML Yellow Paper: Evolutionary AI and Optimisation Studio
Authors:
Lingbo Li,
Leslie Kanthan,
Michail Basios,
Fan Wu,
Manal Adham,
Vitali Avagyan,
Alexis Butler,
Paul Brookes,
Rafail Giavrimis,
Buhong Liu,
Chrystalla Pavlou,
Matthew Truscott,
Vardan Voskanyan
Abstract:
Machine learning model development and optimisation can be a rather cumbersome and resource-intensive process. Custom models are often more difficult to build and deploy, and they require infrastructure and expertise which are often costly to acquire and maintain. Machine learning product development lifecycle must take into account the need to navigate the difficulties of develo** and deploying…
▽ More
Machine learning model development and optimisation can be a rather cumbersome and resource-intensive process. Custom models are often more difficult to build and deploy, and they require infrastructure and expertise which are often costly to acquire and maintain. Machine learning product development lifecycle must take into account the need to navigate the difficulties of develo** and deploying machine learning models. evoML is an AI-powered tool that provides automated functionalities in machine learning model development, optimisation, and model code optimisation. Core functionalities of evoML include data cleaning, exploratory analysis, feature analysis and generation, model optimisation, model evaluation, model code optimisation, and model deployment. Additionally, a key feature of evoML is that it embeds code and model optimisation into the model development process, and includes multi-objective optimisation capabilities.
△ Less
Submitted 20 December, 2022;
originally announced December 2022.
-
Chemical Vapour Deposition Graphene-PMMA Nanolaminates for Flexible Gas Barrier
Authors:
Antonio Baldanza,
Maria Giovanna Pastore Carbone,
Cosimo Brondi,
Anastasios C. Manikas,
Giuseppe Mensitieri,
Christos Pavlou,
Giuseppe Scherillo,
Costas Galiotis
Abstract:
Successful ways of fully exploiting the excellent structural and multifunctional performance of graphene and related materials are of great scientific and technological interest. New opportunities are provided by the fabrication of a novel class of nanocomposites with a nanolaminate architecture. In this work, by using the iterative lift-off/float-on process combined with wet depositions, we incor…
▽ More
Successful ways of fully exploiting the excellent structural and multifunctional performance of graphene and related materials are of great scientific and technological interest. New opportunities are provided by the fabrication of a novel class of nanocomposites with a nanolaminate architecture. In this work, by using the iterative lift-off/float-on process combined with wet depositions, we incorporated cm-size graphene monolayers produced via Chemical Vapour Deposition into a poly (methyl methacrylate) (PMMA) matrix with a controlled, alternate-layered structure. The produced nanolaminate shows a significant improvement in mechanical properties, with enhanced stiffness, strength and toughness, with the addition of only 0.06 vol% of graphene. Furthermore, oxygen and carbon dioxide permeability measurements performed at different relative humidity levels, reveal that the addition of graphene leads to significant reduction of permeability, compared to neat PMMA. Overall, we demonstrate that the produced graphene-PMMA nanolaminate surpasses, in terms of gas barrier properties, the traditional discontinuous graphene-particle composites with a similar filler content. Moreover, we found that the gas permeability through the nanocomposites departs from a monotonic decrease as a function of relative humidity, which is instead evident in the case of the pure PMMA nanolaminate. This work suggests the possible use of Chemical Vapour Deposition graphene-polymer nanolaminates as a flexible gas barrier, thus enlarging the spectrum of applications for this novel material.
△ Less
Submitted 22 August, 2022;
originally announced September 2022.
-
Determination of the elastic moduli of CVD graphene by probing graphene/polymer Bragg stacks
Authors:
Bohai Liu,
Christos Pavlou,
Zuyuan Wang,
Yu Cang,
Costas Galiotis,
George Fytas
Abstract:
Graphene has been widely used in the form of micro-flakes to fabricate composite materials with enhanced mechanical properties. Due to the small size of the inclusions and their random orientation within the matrix, the superior mechanical properties of graphene cannot be fully exploited. Recently, attempts have been made to fabricate nanolaminate composites by interleaving large sheets of chemica…
▽ More
Graphene has been widely used in the form of micro-flakes to fabricate composite materials with enhanced mechanical properties. Due to the small size of the inclusions and their random orientation within the matrix, the superior mechanical properties of graphene cannot be fully exploited. Recently, attempts have been made to fabricate nanolaminate composites by interleaving large sheets of chemical vapor deposition (CVD) monolayer graphene between thin layers of polymer matrices. However, CVD graphene is inevitably accompanied by wrinkles that are formed in the synthesis process, and it remains unknown how the wrinkles affect the mechanical properties of graphene. Here, we employ Brillouin Light Spectroscopy (BLS) to study the elastic moduli of CVD graphene by probing graphene/poly(methylmethacrylate) hybrid Bragg stacks at zero strain. We find the Young's and shear moduli of the CVD graphene, which has wrinkles in the form of sharp elevations of height of about 6 nm and a FWHM of ca. 30 nm, to be 680 and 290 GPa, respectively, with the former being about 30% lower than that of exfoliated, flat graphene. This work sheds light on the elastic properties of CVD graphene and provides a method that can be extended to studying the wrinkle-induced softening effect in other two-dimensional materials.
△ Less
Submitted 27 September, 2021;
originally announced September 2021.
-
Multi-functional 2D hybrid aerogels for gas absorption applications
Authors:
Charalampos Androulidakis,
Maria Kotsidi,
George Gorgolis,
Christos Pavlou,
Labrini Sygellou,
George Paterakis,
Nick Koutroumanis,
Costas Galiotis
Abstract:
Aerogels have attracted significant attention recently due to their ultra-light weight porous structure, mechanical robustness, high electrical conductivity, facile scalability and their use as gas and oil absorbers. Herein, we examine the multi-functional properties of hybrid aerogels consisting of reduced graphene oxide (rGO) integrated with hexagonal boron nitride (hBN) platelets. Using a freez…
▽ More
Aerogels have attracted significant attention recently due to their ultra-light weight porous structure, mechanical robustness, high electrical conductivity, facile scalability and their use as gas and oil absorbers. Herein, we examine the multi-functional properties of hybrid aerogels consisting of reduced graphene oxide (rGO) integrated with hexagonal boron nitride (hBN) platelets. Using a freeze-drying approach, hybrid aerogels are fabricated by simple mixing with various volume fractions of hBN and rGO up to 0.5/0.5 ratio. The fabrication method is simple, cost effective, scalable and can be extended to other 2D materials combinations. The hybrid rGO/hBN aerogels (HAs) are mechanically robust and highly compressible with mechanical properties similar to those of the pure rGO aerogel. We show that the presence of hBN in the HAs enhances the gas absorption capacities of formaldehyde and water vapour up to ~7 and > 8 times, respectively, as compared to pure rGO aerogel. Moreover, the samples show good recoverability, making them highly efficient materials for gas absorption applications and for the protection of artefacts such as paintings in storage facilities. Finally, even in the presence of large quantity of insulating hBN, the HAs are electrically conductive, extending the potential application spectrum of the proposed hybrids to the field of electro-thermal actuators. The work proposed here paves the way for the design and production of novel 2D materials combinations with tailored multi-functionalities suited for a large variety of modern applications.
△ Less
Submitted 27 September, 2021;
originally announced September 2021.
-
Effective EMI shielding behaviour of thin graphene/PMMA nanolaminates in the THz range
Authors:
Christos Pavlou,
Maria Giovanna Pastore Carbone,
Anastasios C Manikas,
George Trakakis,
Can Koral,
Gianpaolo Papari,
Antonello Andreone,
Costas Galiotis
Abstract:
The use of graphene in a form of discontinuous flakes in polymer composites limits the full exploitation of the unique properties of graphene, thus requiring high filler loadings for achieving - for example - satisfactory electrical and mechanical properties. Herein centimetre-scale CVD graphene/polymer nanolaminates have been produced by using an iterative 'lift-off/float-on' process and have bee…
▽ More
The use of graphene in a form of discontinuous flakes in polymer composites limits the full exploitation of the unique properties of graphene, thus requiring high filler loadings for achieving - for example - satisfactory electrical and mechanical properties. Herein centimetre-scale CVD graphene/polymer nanolaminates have been produced by using an iterative 'lift-off/float-on' process and have been found to outperform, for the same graphene content, state-of-the-art flake-based graphene polymer composites in terms of mechanical reinforcement and electrical properties. Most importantly these thin laminate materials show a high electromagnetic interference (EMI) shielding effectiveness, reaching 60 dB for a small thickness of 33 micron, and an absolute EMI shielding effectiveness close to 3* 10 ^5 dB cm ^2 g^-1 which is amongst the highest values for synthetic, non-metallic materials produced to date.
△ Less
Submitted 23 August, 2021;
originally announced August 2021.
-
Mosaic pattern formation in exfoliated graphene by mechanical deformation
Authors:
Maria Giovanna Pastore Carbone,
Anastasios C. Manikas,
Ioanna Souli,
Chistos Pavlou,
Costas Galiotis
Abstract:
Graphene is susceptible to morphological instabilities such as wrinkles and folds, which result from the imposition of thermo-mechanical stresses upon cooling from high temperatures and/ or under biaxial loading. A particular pattern encountered in CVD graphene is that of mosaic formation. Although it is understood that this pattern results from the severe biaxial compression upon cooling from hig…
▽ More
Graphene is susceptible to morphological instabilities such as wrinkles and folds, which result from the imposition of thermo-mechanical stresses upon cooling from high temperatures and/ or under biaxial loading. A particular pattern encountered in CVD graphene is that of mosaic formation. Although it is understood that this pattern results from the severe biaxial compression upon cooling from high temperatures, it has not been possible to create such a complex pattern at room temperature by mechanical loading. Herein, we have managed by means of lateral wrinkling induced by tension and Euler buckling resulting from uniaxial compression upon unloading, to create such patterns in exfoliated graphene. We also show that these patterns can be used as channels for trap** or administering fluids at interstitial space between graphene and its support. This opens a whole dearth of new applications in the area of nano-fluidics but also in photo-electronics and sensor technologies.
△ Less
Submitted 11 April, 2019;
originally announced April 2019.
-
On the Complexity of the Inverse Semivalue Problem for Weighted Voting Games
Authors:
Ilias Diakonikolas,
Chrystalla Pavlou
Abstract:
Weighted voting games are a family of cooperative games, typically used to model voting situations where a number of agents (players) vote against or for a proposal. In such games, a proposal is accepted if an appropriately weighted sum of the votes exceeds a prespecified threshold. As the influence of a player over the voting outcome is not in general proportional to her assigned weight, various…
▽ More
Weighted voting games are a family of cooperative games, typically used to model voting situations where a number of agents (players) vote against or for a proposal. In such games, a proposal is accepted if an appropriately weighted sum of the votes exceeds a prespecified threshold. As the influence of a player over the voting outcome is not in general proportional to her assigned weight, various power indices have been proposed to measure each player's influence. The inverse power index problem is the problem of designing a weighted voting game that achieves a set of target influences according to a predefined power index. In this work, we study the computational complexity of the inverse problem when the power index belongs to the class of semivalues. We prove that the inverse problem is computationally intractable for a broad family of semivalues, including all regular semivalues. As a special case of our general result, we establish computational hardness of the inverse problem for the Banzhaf indices and the Shapley values, arguably the most popular power indices.
△ Less
Submitted 31 December, 2018;
originally announced December 2018.
