Acquiring reusable skills in intrinsically motivated reinforcement learning

Journal of Intelligent Manufacturing - This paper proposes a novel incremental model for acquiring skills and using them in Intrinsically Motivated Reinforcement Learning (IMRL). In this model, the...     

Effects of peroxide and phenolic cure systems on characteristics of the filled ethylene–propylene–diene monomer rubber (EPDM)

The effects of three curing systems, peroxide, peroxide–phenolic combination, and phenolic on selected properties of cured carbon black‐filled ethylene–propylene–diene monomer rubber (EPDM) were investigated. The cured rubbers immersed in hot amine solution to evaluate their suitability for seal and gasket industry at elevated temperature and amine environments. These tests were essential for evaluating the durability of the gasket in a gas refinery. The Fourier transform infrared spectroscopy spectrums revealed that the phenolic crosslink was constructed between rubber macromolecules during the curing process. The changing curing system from peroxide to peroxide–phenolic and phenolic increased the glass transition temperature of the filled cured rubbers between 3 and 5 °C. There was not any significant difference between thermogravimetric analysis thermographs of the selected cured rubbers with various cure systems and the residues ranged between 45% and 47%. Unlike of peroxide curing system, a dual phase was observed from scanning electron microscopy micrographs for peroxide–phenolic and phenolic cure systems. The phenolic cure system was not beneficial for rubber curing although, it reduced scorch time of the curing process. For the most studied mechanical properties, phenolic cure system deteriorated mechanical properties for both, aged and unaged cured rubbers. Increasing the amount of diene monomer in EPDM structure was beneficial for phenolic rubber cure system. © 2018 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2018 , 135, 46213.     

Multi‐“Color” Delineation of Bone Microdamages Using Ligand‐Directed Sub‐5 nm Hafnia Nanodots and Photon Counting CT Imaging

The early detection of bone microdamages is crucial to make informed decisions about the therapy and taking precautionary treatments to avoid catastrophic fractures. Conventional computed tomography (CT) imaging faces obstacles in detecting bone microdamages due to the strong self‐attenuation of photons from bone and poor spatial resolution. Recent advances in CT technology as well as novel imaging probes can address this problem effectively. Herein, the bone microdamage imaging is demonstrated using ligand‐directed nanoparticles in conjunction with photon counting spectral CT. For the first time, Gram‐scale synthesis of hafnia (HfO₂) nanoparticles is reported with surface modification by a chelator moiety. The feasibility of delineating these nanoparticles from bone and soft tissue of muscle is demonstrated with photon counting spectral CT equipped with advanced detector technology. The ex vivo and in vivo studies point to the accumulation of hafnia nanoparticles at microdamage site featuring distinct spectral signal. Due to their small sub‐5 nm size, hafnia nanoparticles are excreted through reticuloendothelial system organs without noticeable aggregation while not triggering any adverse side effects based on histological and liver enzyme function assessments. These preclinical studies highlight the potential of HfO₂‐based nanoparticle contrast agents for skeletal system diseases due to their well‐placed K‐edge binding energy.     

Representation and analysis of enterprise models with semantic techniques: an application to ArchiMate,e3value and business model canvas

Enterprise models assist the governance and transformation of organizations through the specification, communication and analysis of strategy, goals, processes, information, along with the underlying application and technological infrastructure. Such models cross-cut different concerns and are often conceptualized using domain-specific modelling languages. This paper explores the application of graph-based semantic techniques to specify, integrate and analyse multiple, heterogeneous enterprise models. In particular, the proposal described in this paper (1) specifies enterprise models as ontological schemas, (2) uses transformation mapping functions to integrate the ontological schemas and (3) analyses the integrated schemas with graph querying and logical inference. The proposal is evaluated through a scenario that integrates three distinct enterprise modelling languages: the business model canvas, e3value, and the business layer of the ArchiMate language. The results show, on the one hand, that the graph-based approach is able to handle the specification, integration and analysis of enterprise models represented with different modelling languages and, on the other, that the integration challenge resides in defining appropriate mapping functions between the schemas.     

Design of X‐band Moreno cross‐guide coupler based on superformula curves

This article deals with the design of Moreno cross‐guide couplers based on supershapes for X‐band applications. Crossed‐waveguide couplers are mainly used due to their compact structures. In these couplers, cross‐aperture structures are usually employed to offer flat coupling and high isolation. In the present article, the possible shapes for apertures and metal inserts that can be derived by the superfomula curves are explored and the effects of variations of superformula parameters are investigated on the performance of Moreno coupler. Finally, the proposed Moreno coupler is validated through fabrication and measurement. The experimental validation shows an excellent agreement with the simulated results. In the frequency range from 8 to 12.5 GHz, the measured coupling value changes from 18.8 to 20.8 dB and the directivity is better than 38 dB and 29 dB from 8 to 11 GHz and 11 to 12.5 GHz, respectively. The results are valuable for the design and evaluation of broadband high directive waveguide couplers.     

Synthesis,characterization, and optimization of α‐Fe2O3/silica nanocomposites using homogenous precipitation method

The purpose of this study was to systematically synthesize and characterize the high surface area 10 wt% nanocomposites of α‐Fe₂O₃ (hematite)/silica using a simple and economically effective homogenous precipitation (HP) route via Response Surface Method combined with Central Composite Design (CCD). Accordingly, the RSM‐CCD approach including 20 experiments was designed to investigate the effects of three factors including concentration of iron chloride solution, pH and calcinations temperature on the final surface area of α‐Fe₂O₃/silica nanocomposites. The optimum surface area was 373 m²/g at the condition including iron chloride concentration of 0.018 mol/L, pH=8.95, and calcination temperature of 573°C.     

A Scenario-Based Management of Water Resources and Supply Systems Using a Combined System Dynamics and Compromise Programming Approach

Water Resources Management - Long-term sustainability in water supply systems is a major challenge due to water resources depletion, climate change and population growth. This paper presents a...     

Depression detection from sMRI and rs-fMRI images using machine learning

Major Depression Disorder (MDD) is a common mental disorder that negatively affects many people’s lives worldwide. Developing an automated method to find useful diagnostic biomarkers from brain imaging data would help clinicians to detect MDD in its early stages. Depression is known to be a brain connectivity disorder problem. In this paper, we present a brain connectivity-based machine learning (ML) workflow that utilizes similarity/dissimilarity of spatial cubes in brain MRI images as features for depression detection. The proposed workflow provides a unified framework applicable to both structural MRI images and resting-state functional MRI images. Several cube similarity measures have been explored, including Pearson or Spearman correlations, Minimum Distance Covariance, or inverse of Minimum Distance Covariance. Discriminative features from the cube similarity matrix are chosen with the Wilcoxon rank-sum test. The extracted features are fed into machine learning classifiers to train MDD prediction models. To address the challenge of data imbalance in MDD detection, oversampling is performed to balance the training data. The proposed workflow is evaluated through experiments on three independent public datasets, all imbalanced, of structural MRI and resting-state fMRI images with depression labels. Experimental results show good performance on all three datasets in terms of prediction accuracy, specificity, sensitivity, and area under the Receiver Operating Characteristic (ROC) curve. The use of features from both structured MRI and resting state functional MRI is also investigated.