Hybrid Version Research Articles

An Overview of the Application of Harmony Search for Chemical Engineering Optimization

Harmony search algorithm and its variants have been used in several applications in medicine, telecommunications, computer science, and engineering. This article reviews the global and multi-objective optimization for chemical engineering using harmony search. The main features of the HS method and several of its popular variants and hybrid versions including their relevant algorithm characteristics are described and discussed. A variety of global and multi-objective optimization problems from chemical engineering and their resolution using HS-based methods are also included. These problems involve thermodynamic calculations (phase stability analysis, phase equilibrium calculations, parameter estimation, and azeotrope calculation), heat exchanger design, distillation simulation, life cycle analysis, and water distribution systems, among others. Remarks on future developments of HS and its related algorithms for global and multi-objective optimization in chemical engineering are also provided in this review. HS is a reliable and promising stochastic optimizer to resolve challenging global and multi-objective optimization problems for process systems engineering.

Multi-class motion-based semantic segmentation for ureteroscopy and laser lithotripsy.

Ureteroscopy with laser lithotripsy has evolved as the most commonly used technique for the treatment of kidney stones. Automated segmentation of kidney stones and the laser fiber is an essential initial step to performing any automated quantitative analysis, particularly stone-size estimation, that can be used by the surgeon to decide if the stone requires further fragmentation. However, factors such as turbid fluid inside the cavity, specularities, motion blur due to kidney movements and camera motion, bleeding, and stone debris impact the quality of vision within the kidney, leading to extended operative times. To the best of our knowledge, this is the first attempt made towards multi-class segmentation in ureteroscopy and laser lithotripsy data. We propose an end-to-end convolution neural network (CNN) based learning framework for the segmentation of stones and laser fiber. The proposed approach utilizes two sub-networks: (I) HybResUNet, a hybrid version of residual U-Net, that uses residual connections in the encoder path of the U-Net to improve semantic predictions, and (II) a DVFNet that generates deformation vector field (DVF) predictions by leveraging motion differences between the adjacent video frames which is then used to prune the prediction maps. We also present ablation studies that combine different dilated convolutions, recurrent and residual connections, atrous spatial pyramid pooling, and attention gate models. Further, we propose a compound loss function that significantly boosts the segmentation performance in our data. We have also provided an ablation study to determine the optimal data augmentation strategy for our dataset. Our qualitative and quantitative results illustrate that our proposed method outperforms state-of-the-art methods such as UNet and DeepLabv3+ showing a DSC improvement of 4.15% and 13.34%, respectively, in our in vivo test dataset. We further show that our proposed model outperforms state-of-the-art methods on an unseen out-of-sample clinical dataset with a DSC improvement of 9.61%, 11%, and 5.24% over UNet, HybResUNet, and DeepLabv3+, respectively in the case of the stone class and an improvement of 31.79%, 22.15%, and 10.42% over UNet, HybResUNet, and DeepLabv3+, respectively, in case of the laser class.

Performance enhancement of PRNU-based source identification for smart video surveillance

This paper introduces a current signal-based source verification (SSV) system for images in video surveillance networks including the cloud. Using a signal, the well-known photo-response non-uniformity (PRNU) can be used, which is unique and intrinsic in every digital image taken by a source camera, like fingerprints. The SSV system using PRNU has proved before to be useful for reliable video source identification in both network- and cloud-based video surveillance. However, in the era of smart living, security video systems have become part of the IoT devices which typically have limited resources such as low computation, power, storage and memory. To address these problems in the IoT applications, the effects of I-frames only and infra-red night scenes are studied as well as two proposed approaches for the SSV system. Then a hybrid version of the SSV scheme is further suggested, in combination with the best approach using averaged noise residues (for reduced false positive rate), and a recent technique using spatial domain averaged frames (for reduced computational complexity). The enhanced performance of the improved SSV system for smart video surveillance has been verified through tests.

Chronic Kidney Disease Prognostication Utilizing Hybrid Ensemble Learning

The chronic kidney failure is a extreme fitness issue and if not detected and handled on the early degrees, it can be very lethal. Hence the essential goal of this paper is to increase a dependable system learning version which predicts the CKD with a high accuracy price. The CKD statistics set is downloaded from the famous UCI ML repository but it suffers from a variety of lacking values. To deal with the lacking values KNN Imputation is used. Feature selection is likewise completed with the assist of information benefit because the dataset is massive and consequently the fee of modelling can be very excessive. Various other pre-processing steps like label encoding and Min-max normalization is executed to obtain a clean dataset. After pre-processing, diverse ML algorithms like logistic regression, naïve bayes, synthetic neural network and random wooded area are implemented and their performances are in comparison with the assist of diverse overall performance metrics. A hybrid of Random Forest and Adaboost algorithm is proposed and it achieves a higher accuracy when in comparison to the opposite person element models and subsequently it may be proved that the proposed hybrid version is an awful lot better and accurate in diagnosing CKD. goal of this paper is to increase a dependable system learning version which predicts the CKD with a high accuracy price. The CKD statistics set is downloaded from the famous UCI ML repository but it suffers from a variety of lacking values. To deal with the lacking values KNN Imputation is used. Feature selection is likewise completed with the assist of information benefit because the dataset is massive and consequently the fee of modelling can be very excessive. Various other pre-processing steps like label encoding and Min-max normalization is executed to obtain a clean dataset. After pre-processing, diverse ML algorithms like logistic regression, naïve bayes, synthetic neural network and random wooded area are implemented and their performances are in comparison with the assist of diverse overall performance metrics. A hybrid of Random Forest and Adaboost algorithm is proposed and it achieves

Applying autonomous hybrid agent-based computing to difficult optimization problems

Evolutionary multi-agent systems (EMASs) are very good at dealing with difficult, multi-dimensional problems, their efficacy was proven theoretically based on analysis of the relevant Markov-Chain based model. Now the research continues on introducing autonomous hybridization into EMAS. This paper focuses on a proposed hybrid version of the EMAS, and covers selection and introduction of a number of hybrid operators and defining rules for starting the hybrid steps of the main algorithm. Those hybrid steps leverage existing, well-known and proven to be efficient metaheuristics, and integrate their results into the main algorithm. The discussed modifications are evaluated based on a number of difficult continuous-optimization benchmarks.

Marine predator inspired naked mole-rat algorithm for global optimization

This paper proposes a hybrid version of marine predator algorithm (MPA) and naked mole-rat algorithm (NMRA) to aggregate the strengths of both algorithms. The new proposed algorithm is named as MpNMRA and designed to overcome the inherent drawbacks of MPA (slow exploitation) and NMRA (limited exploration). The algorithm adds the basic structure of MPA to the worker phase of NMRA, while keeping all the major parameters of both the algorithm. The major parameters of both the algorithms are subjected to four different mutation strategies namely, exponential, linear, simulated annealing and logarithmic mutation strategies. The concept of simulated annealing-based mutation is found to be best for most of the parameters, whereas in some cases exponentially decreasing weights provide better results. Leveraging on the best mutation strategies for all the parameters, the proposed MpNMRA is tested on CEC2005, CEC2014 and CEC 2019 benchmark problems. The experimental results demonstrate that MpNMRA provides best results when compared to other algorithms in the literature on higher-dimensional problems. This work also considers solving three real-world optimization problems and training of a multi-layer perceptron using the proposed algorithm. Statistical results obtained from Wilcoxon’s rank-sum test, Freidman’s test and computational complexity further proves that the proposed algorithm is highly efficient and provide superior results.11The link of source-code will be added after publishing.

Stochastically switching diffusion with partially reactive surfaces.

In this paper we develop a hybrid version of the encounter-based approach to diffusion-mediated absorption at a reactive surface, which takes into account stochastic switching of a diffusing particle's conformational state. For simplicity, we consider a two-state model in which the probability of surface absorption depends on the current particle state and the amount of time the particle has spent in a neighborhood of the surface in each state. The latter is determined by a pair of local times ℓ_{n,t}, n=0,1, which are Brownian functionals that keep track of particle-surface encounters over the time interval [0,t]. We proceed by constructing a differential Chapman-Kolmogorov equationfor a pair of generalized propagators P_{n}(x,ℓ_{0},ℓ_{1},t), where P_{n} is the joint probability density for the set (X_{t},ℓ_{0,t},ℓ_{1,t}) when N_{t}=n, where X_{t} denotes the particle position and N_{t} is the corresponding conformational state. Performing a double Laplace transform with respect to ℓ_{0},ℓ_{1} yields an effective system of equationsdescribing diffusion in a bounded domain Ω, in which there is switching between two Robin boundary conditions on ∂Ω. The corresponding constant reactivities are κ_{j}=Dz_{j} and j=0,1, where z_{j} is the Laplace variable corresponding to ℓ_{j} and D is the diffusivity. Given the solution for the propagators in Laplace space, we construct a corresponding probabilistic model for partial absorption, which requires finding the inverse Laplace transform with respect to z_{0},z_{1}. We illustrate the theory by considering diffusion of a particle on the half-line with the boundary at x=0 effectively switching between a totally reflecting and a partially absorbing state. We calculate the flux due to absorption and use this to compute the resulting MFPT in the presence of a renewal-based stochastic resetting protocol. The latter resets the position and conformational state of the particle as well as the corresponding local times. Finally, we indicate how to extend the analysis to higher spatial dimensions using the spectral theory of Dirichlet-to-Neumann operators.

인스타그램기반 이미지딥러닝을 활용한 개인 맞춤형 미술작품추천 방법

This research suggests a way of recommendation criteria for shareholders’ style in artwork using Image Deep learning. In terms of practices, visitors to MoMA (Museum of Modern Art) can only be limited as participants. In this aspect, using online data of the person and artwork is access to the concept of interaction as a subject. In this paper, each exhibition explored an user-centered perspective and implementation methods were explored by a person’s data. The data of Person and artwork image have been made on Instagram. Scraping and retraining image posts of Likes and Posting - two folders were created as a label - with a targeted person’s Instagram data. Bottleneck, the final phase of convolutional neural network fused using Tensor Flow for auto image classification. Then picture images of the MoMA were tested. The test result shows the consistency of labels. Three standards of recommendation demonstrate artwork image for a person by a person. first, personal content ‘Posting’, second preference on other contents ‘Like’, Last but not least, hybrid both to export with the highest consistency ‘Posting Score 1st’ and Likes score 1st. This hybrid version of data is expected to use this process of research has implications for the shareholders that represent functions of access to various visual cultures.

Moth Flame Optimization: Theory, Modifications, Hybridizations, and Applications.

The Moth flame optimization (MFO) algorithm belongs to the swarm intelligence family and is applied to solve complex real-world optimization problems in numerous domains. MFO and its variants are easy to understand and simple to operate. However, these algorithms have successfully solved optimization problems in different areas such as power and energy systems, engineering design, economic dispatch, image processing, and medical applications. A comprehensive review of MFO variants is presented in this context, including the classic version, binary types, modified versions, hybrid versions, multi-objective versions, and application part of the MFO algorithm in various sectors. Finally, the evaluation of the MFO algorithm is presented to measure its performance compared to other algorithms. The main focus of this literature is to present a survey and review the MFO and its applications. Also, the concluding remark section discusses some possible future research directions of the MFO algorithm and its variants.

A hybrid multi-reference subband control strategy for active noise control headphones

Multi-reference feedforward Active Noise Control (ANC) has been shown to provide increased noise mitigation capability in ANC headphones compared to a single reference feedforward controllers, due to the availability of additional time advanced information. Recent studies have also shown that the integration of a single-reference feedforward controller and a feedback controller increases both the noise cancellation bandwidth and system stability, therefore, a corresponding hybrid version of the multi-reference control strategy has been developed and is presented in this paper. The hybrid multi-reference ANC (HMRANC) system is comprised of a multi-reference feedforward control unit acting in conjunction with a feedback ANC system. In addition, to overcome the high computational demand of the hybrid structure, a delayless subband version of the HMRANC system is also presented in this paper. The proposed HMRANC and delayless subband HMRANC headphone systems achieve an improvement of 5–10 dB over the previously proposed hybrid ANC headphone system for different acoustic field scenarios. Moreover, the delayless subband HMRANC headphone system is 40% less computationally demanding then HMRANC and 10% less demanding than HANC headphone systems.

Design and area optimization of CMOS operational amplifier circuit using hybrid flower pollination algorithm for IoT end-node devices

CMOS operational amplifiers are widely used in internet of things (IoT) systems due to their excellent performance at low input voltage. However, the design and optimization of high-performance operational amplifiers are time-consuming processes, even for experienced analog circuit designers. For circuit designer, it is also difficult to quickly evaluate the feasibility of a given specification in a technology node. This paper presents an optimal sizing of operational amplifiers based on the swarm intelligence method. A hybrid version of the Flower Pollination Algorithm (HFPA) was introduced to solve an analog circuit transistor sizing problem effectively and also efficiently reduce the design search space. The proposed method was employed to reduce the transistor size of an operational amplifier to fit for end-node IoT devices. The simulation results show that the HFPA algorithm improved circuit performance in respect of area, power dissipation, and gain compared with other two existing methods. In addition to simulation results, spice validation through cadence is also employed to validate the specifications of the proposed circuit. Therefore the proposed amplifier design is mostly suitable of IoT end-node devices.

Deep null fixing on optimal compromise among sum and difference patterns of thinned arrays through the Schelkunoff unit circle representation

Difference far-field patterns represent a way for pin-pointing a target in both azimuth and elevation, extremely useful in radar applications. At the present work, an innovative method for synthesizing good compromise solutions among sum and difference patterns providing low complexity of the antenna feeding network for uniform thinned arrays is addressed. This procedure uses a hybrid version of the Simulated Annealing algorithm (hybrid SA) to optimize a cost function of radiation characteristics for both sum and difference patterns as peak directivity and side lobe level (SLL) while fixing deep nulls. In this framework, examples of half-wavelength spaced linear arrays from 40 to 120 elements were analyzed, as well as an extension to planar arrays by means of separable distributions was developed. The performance of the method is analyzed with different examples and its potential outlined, showing the ability of fixing deep nulls in both sum and difference patterns which share the same uniform excitation relative amplitudes.

The Hardy–Littlewood–Chowla conjecture in the presence of a Siegel zero

Assuming that Siegel zeros exist, we prove a hybrid version of the Chowla and Hardy–Littlewood prime tuples conjectures. Thus, for an infinite sequence of natural numbers x $x$ , and any distinct integers h 1 , ⋯ , h k , h 1 ′ , ⋯ , h ℓ ′ $h_1,\dots ,h_k,h^{\prime }_1,\dots ,h^{\prime }_\ell$ , we establish an asymptotic formula for ∑ n ⩽ x Λ ( n + h 1 ) ⋯ Λ ( n + h k ) λ ( n + h 1 ′ ) ⋯ λ ( n + h ℓ ′ ) \begin{equation*} \hspace*{13pt}\sum _{n\leqslant x}\Lambda (n+h_1)\cdots \Lambda (n+h_k)\lambda (n+h_{1}^{\prime })\cdots \lambda (n+h_{\ell }^{\prime })\hspace*{-13pt} \end{equation*} for any 0 ⩽ k ⩽ 2 $0\leqslant k\leqslant 2$ and ℓ ⩾ 0 $\ell \geqslant 0$ . Specializing to either ℓ = 0 $\ell =0$ or k = 0 $k=0$ , we deduce the previously known results on the Hardy–Littlewood (or twin primes) conjecture and the Chowla conjecture under the existence of Siegel zeros, due to Heath-Brown and Chinis, respectively. The range of validity of our asymptotic formula is wider than in these previous results.

(Invited, Digital Presentation) Synthesis of Multi-Metallic Clusters Using a Dendrimer Reactor

Dendrimers are highly branched organic macromolecules with successive layers or “generations” of branch units surrounding a central core. Organic inorganic hybrid versions have also been produced, by trapping metal ions or metal clusters within the voids of the dendrimers. Their unusual, tree-like topology endows these nanometer-sized macromolecules with a gradient in branch density from the interior to the exterior, which can be exploited to direct the transfer of charge and energy from the dendrimer periphery to its core.We show that tinchloride, SnCl2 and FeCl3 complexes to the imines groups of a spherical polyphenylazomethine dendrimer in a stepwise fashion according to an electron gradient, with complexation in a more peripheral generation proceeding only after complexation in generations closer to the core has been completed. By attaching an electron-withdrawing group to the dendrimer core, we are able to change the complexation pattern, so that the core imines are complexed last. By further extending this strategy, it should be possible to control the number and location of metal ions incorporated into dendrimer structures, which might and uses as tailored catalysts, building blocks, or fine-controlled clusters for advanced materials. The metal-assembly in a discrete molecule can be converted to a size-regulated metal particle with a size smaller than 1 nm as a molecular reactor. Due to the well-defined number of metal clusters in the subnanometer size region, its property is much different from that of bulk or general metal nanoparticles. The chemistry of metal clusters on the sub-nanometer scale is not yet well understood because metal clusters, especially multi-metallic clusters, are difficult to synthesize with control over size and composition. The template synthesis of multi-metallic sub-nanoclusters is achieved using a phenylazomethine dendrimer as a macromolecular template.

COMPSAC 2022: A Retrospective

Welcome to my annual retrospective on COMPSAC, which, unfortunately, was virtual again. Despite that, or maybe because of it, the last few months before the conference were very busy for our volunteers. Complicating matters was the difficulty, early this year, trying to forecast the movement of the Coronavirus and its likely effect on COMPSAC's attendance in a “real” environment. In April, our Standing Committee decided to take a very conservative approach and go virtual again rather than opt for an in-person, or even a hybrid version of the conference. With surges from evolving virus variants constantly reported worldwide, there was no reason to regret that decision.

An Intelligent Agent Used to Predict the Queer Changes in the Prices of Various Commodities

Abstract: Commodity costs and the prediction of the remarkable adjustments is a tough mission in oil, and gold-generating nations along with UAE, USA in addition to borrowing nations along with India. It is beneficial for each character to music the costs earlier than the unanticipated ascents and descents as we're going to encompass many elements which can also additionally replicate the adjustments in costs along with the conflicts among the nations, new standards or rules being introduced, authorities adjustments, ideology adjustments, and the conflict among nations (eg. Russia Ukraine War which led to a 30 percentage increment in crude oil costs), etc. Predicting those costs is amongst one of the maximum complicated classes to investigate due to the fact fluctuations withinside the charge of those commodities are particularly irregular, nonlinear, and range dynamically with excessive uncertainty. This innovation is proposed as a hybrid version for herbal fueloline, crude oil, and gold charge prediction on a unmarried platform that makes use of complicated statistical, predictive analytics, and superior deep studying algorithms. We validate the end result and examine those empirical outcomes with different studies withinside the literature. The proposed innovation could have better accuracy that is extra strong and reliable. We have many regions to goal withinside the industry, in which inventory marketplace corporations are one of the maximum essential clients, and on the alternative hand, the folks that need to music the records and who're interested by extracting the results in order that it addresses, train the humans which allows in knocking down the inflation charge, authorities corporations that may are expecting the adjustments similarly earlier than can paintings on vital steps to govern those adjustments.

Improved proton-transfer barriers with van der Waals density functionals: Role of repulsive non-local correlation.

Proton-transfer (PT) between organic complexes is a common and important biochemical process. Unfortunately, PT energy barriers are difficult to accurately predict using density functional theory (DFT); in particular, using the generalized gradient approximation (GGA) tends to underestimate PT barriers. Moreover, PT typically occurs in environments where dispersion forces contribute to the cohesion of the system; thus, a suitable exchange-correlation functional should accurately describe both dispersion forces and PT barriers. This paper provides benchmark results for the PT barriers of several density functionals, including several variants of the van der Waals density functional (vdW-DF). The benchmark set comprises small organic molecules with inter- and intra-molecular PT. The results show that replacing GGA correlation with a fully non-local vdW-DF correlation increases the PT barriers, making it closer to the quantum chemical reference values. In contrast, including non-local correlations with the Vydrov-Voorhis method or dispersion-corrections at the DFT-D3 or the Tkatchenko-Scheffler level has barely any impact on the PT barriers. Hybrid functionals also increase and improve the energies, resulting in an excellent performance of hybrid versions of vdW-DF-cx and vdW-DF2-B86R. For the formic acid dimer PT system, we analyzed the GGA exchange and non-local correlation contributions. The analysis shows that the repulsive part of the non-local correlation kernel plays a key role in the PT energy barriers predicted with vdW-DF.

Inflation Dynamics in the Czech Republic: New Evidence on the Cost Based Hybrid New Keynesian Phillips Curve

New Keynesian form of the Phillips Curve assumes a shortterm trade-off between inflation and real economic activity, either in unit labour cost or output gap specification. Extending the New Keynesian Phillips Curve (NKPC) with a backward-looking price setting, our aim is to examine the impact of the inflation expectations, lagged inflation, unit labour cost, import prices and real effective exchange rate on the inflation dynamics in the Czech Republic between 2000M1 and 2020M12. Dealing with non-stationary and cointegrated time series, we opted for an Error Correction Model (ECM) and an Autoregressive Distributed Lag (ARDL) model with the variables integrated of order I(1). ARDL model in differences was also compared with an ARDL model in levels. The main findings of our analysis can be summarized as follows. Firstly, our analysis indicates that the ARDL model using non-stationary time series generates spurious regression results. Secondly, the results from the ECM model and the ARDL model with first order differencing find the inflation expectations and unit labour cost statistically significant, confirming the existence of a small open economy NKPC for the Czech Republic. Thirdly, even though the backward-looking inflation setting represents a cornerstone of the hybrid NKPC, we have found no evidence for lagged inflation to have an impact on the inflation dynamics in the Czech Republic between 2000M1 and 2020M12. We thereby reject the hybrid version of the NKPC. Lastly, in spite of the Czech Republic representing a small open economy, we observed weak statistical evidence suggesting that external factors (exchange rate, import prices) influence the inflation dynamics in the Czech Republic.

A hybrid Dantzig-Wolfe decomposition algorithm for the multi-floor facility layout problem

The multi-floor facility layout problem (MFLP) is one of the most important and complex facility layout problems that has many applications in designing the facilities of manufacturing and service sectors. In this study, a hybrid version of the Dantzig-Wolfe decomposition algorithm is proposed to solve the MFLP for the first time. The proposed solution approach is performed in two steps. In the first step, a mathematical formulation is applied to assign the departments to the floors in a way that the departments with higher material flow between them be located on the same or closer floors. In the second step, the output of the first step is considered and the MFLP is decomposed into a master problem and some sub-problems to form the Dantzig-Wolfe decomposition algorithm and find the optimal layout of each floor separately. Then the integrated layout of multiple floors is formed easily. The proposed algorithm is evaluated using some sample problems from the literature and some newly generated test problems. The obtained results show the superiority of the proposed algorithm compared to the approaches of the literature.

Comments on: hybrid semiparametric Bayesian networks

This note comments on the article of David Atienza, Pedro Larrañaga and Concha Bielza in which they first review recent contributions to Bayesian networks and then introduce a new hybrid version. It combines parametric and nonparametric density estimates for continuous variables by simultaneously allowing for discrete parents. We discuss straightforward extensions of the linear Gaussian parts and potential smoothing over the outcomes of discrete parents and conclude with some minor comments.