Evolutionary design of neural network architectures: a review of three decades of research

We present a comprehensive review of the evolutionary design of neural network architectures. This work is motivated by the fact that the success of an Artificial Neural Network (ANN) highly depends on its architecture and among many approaches Evolutionary Computation, which is a set of global-search methods inspired by biological evolution has been proved to be an efficient approach for optimizing neural network structures. Initial attempts for automating architecture design by applying evolutionary approaches start in the late 1980s and have attracted significant interest until today. In this context, we examined the historical progress and analyzed all relevant scientific papers with a special emphasis on how evolutionary computation techniques were adopted and various encoding strategies proposed. We summarized key aspects of methodology, discussed common challenges, and investigated the works in chronological order by dividing the entire timeframe into three periods. The first period covers early works focusing on the optimization of simple ANN architectures with a variety of solutions proposed on chromosome representation. In the second period, the rise of more powerful methods and hybrid approaches were surveyed. In parallel with the recent advances, the last period covers the Deep Learning Era, in which research direction is shifted towards configuring advanced models of deep neural networks. Finally, we propose open problems for future research in the field of neural architecture search and provide insights for fully automated machine learning. Our aim is to provide a complete reference of works in this subject and guide researchers towards promising directions.

     

Online tax payment systems as an emergent aspect of governmental transformation

The achievement of governmental transformation through the use of electronically delivered services is a worthy goal that requires significant planning and research to achieve. In order to reach transformational paradigm shifts in governmental operation, it will first be necessary to understand and optimize present governmental e-Service provisions. Of these, the revenue function of taxation is paramount. This paper describes factors related to the use and acceptance by accounting professionals of information technology intended to facilitate electronic tax filing systems. Though tested in the context of governmental tax management systems in Turkey, our findings on the use and acceptance of e-Tax systems are relevant and applicable to a great number of nations and contexts as the ongoing electronic transformation of the governmental revenue system contributes to efforts to transform governments through alternative services delivery venues and channels. We discover that intention to use automated systems as part of the governmental treasury function transformation is hindered by factors that mediate actual plans to do so, mostly in terms of normative pressures and perceptions of behavioral control, which training and education may well ameliorate. Hence, transformation of the treasury function in Turkey is only partially complete and will require additional support, direction and training on the part of the government in its interactions with the tax professionals who interact with the emergent automated system.     

复杂度直方图及其在图像检索中的应用

该文提出了一种新的描述图像纹理信息的复杂度直方图特征,实验结果表明该特征用于图像检索可以取得较好的检索效果。文章同时将该特征与图像色彩直方图组合特征用于图像检索,实验结果证实使用组合特征可以得到更高的检索效率。     

Halogen lamp–microwave combination baking of cookies

The effects of halogen lamp–microwave combination baking on the quality of cookies in terms of texture, color and spread ratio were studied. In addition, gelatinization and pasting properties of cookies baked in different ovens were assessed by using a Rapid Visco Analyzer. The hardness values of the cookies increased with increasing baking time and/or halogen power. The microwave power also contributed to cookie hardness. The lightness values of the cookies decreased, while the a* and E values increased linearly during baking at different halogen lamp and microwave powers. The spread ratio of conventionally baked cookies was significantly lower than that of the other cookies. Halogen lamp power, microwave power and baking time were found to have a significant effect on the spread ratio of the cookies. The moisture contents of the cookies decreased during baking as the halogen lamp and/or the microwave power level increased. The best baking condition in a halogen lamp–microwave combination oven to produce cookies having similar quality parameters as conventionally baked ones was baking at 70% halogen lamp and 20% microwave power levels for 5.5 min. The baking time of these cookies is half of that required in conventional baking.     

Pressure‐dependent Raman modes near the cubic‐tetragonal transition in strontium titanate

The pressure dependence of the Raman frequency shifts of various Raman modes is calculated at room temperature using the volume data from the literature for the cubic‐tetragonal transition in SrTiO₃. The isothermal mode Grüneisen parameters of those Raman modes are obtained, which decrease with increasing pressure for this molecular crystal. Calculated Raman frequencies are then used to predict the damping constant and the inverse relaxation time of those Raman modes as a function of pressure by means of the pseudospin‐phonon (PS) coupled model and the energy fluctuation (EF) model to describe the cubic‐tetragonal transition in SrTiO₃. Also, the values of the activation energy are extracted for the Raman modes studied using both models (PS and EF). Our predicted damping constant and the inverse relaxation time for the Raman modes, can be compared with the experimental measurements close to the cubic‐tetragonal transition in SrTiO₃.     

Optimum design of rotor-bearing system stability performance comparing an evolutionary algorithm versus a conventional method

The main intent of this paper is to formulate, demonstrate, and validate a practical means of implementing an evolutionary optimization technique in a rotor-bearing system. The optimum design of a flexible rotor supported on three-lobe bearings is studied for the optimal performance considering system stability along with other design criteria such as fluid film thickness, power loss, film temperature, and film pressure using the genetic algorithm and the method of feasible directions. The results for different operating speed values obtained and presented in this study are to provide a comparison of these two methods, and to show the potential of the genetic algorithm in optimization of rotor-bearing systems. The genetic algorithm obtained reasonably good results for the objective function and comparable to the method of feasible directions. Thus, the genetic algorithm provides the designer an alternative design optimization approach for rotor-bearing systems.     

Effects of suni-bug (<Emphasis Type="Italic">Eurygaster</Emphasis> spp.) damage on size distribution of durum wheat (<Emphasis Type="Italic">Triticum durum</Emphasis> L.) proteins

Wholemeal samples were obtained from five durum wheat cultivars at two different bug (Eurygaster spp.) damage levels (medium and high damage). The samples were incubated (60 and 120 min) and used in size exclusion high performance liquid chromatography (SE-HPLC) analyses. The results showed that the amount of larger polymeric protein (TP1) and smaller polymeric protein (TP2) obtained from total (sodium dodecyl sulfate soluble) proteins decreased significantly in the bug-damaged samples, while the amount of total larger monomeric proteins (TP3) increased. The polymeric/monomeric protein ratio of all cultivars decreased at 60 min of incubation with increasing damage level. For all cultivars, the ratio of unextractable polymeric protein (UPP%) significantly decreased at 60 min of incubation except cv. Diyarbakir. The results suggested that bug protease caused depolymerization and/or disaggregation of polymeric proteins to lower their average molecular size. The changes in protein structure as determined using SE-HPLC supported by the decreases in gluten content and gluten index values which decreased with suni-bug damage. Deteriorative effects of bug damage on durum wheat quality were found to be quite similar to those on bread wheats.     

Oxidative DNA damage correlates with carotid artery atherosclerosis in hemodialysis patients

Oxidative stress is accepted as a nonclassical cardiovascular risk factor in chronic renal failure patients. The aim of this study was to evaluate the relation between oxidative DNA damage (8‐hydroxy‐2′‐deoxyguanosine/deoxyguanosine [8‐OHdG/dG] ratio), oxidative stress biomarkers, antioxidant enzymes, and carotid artery intima‐media thickness (CIMT) in hemodialysis (HD) patients. Forty chronic HD patients without known atherosclerotic disease and 48 age‐ and sex‐matched healthy individuals were included in the study. Plasma malondialdehyde (MDA) levels and 8‐OHdG/dG ratio were determined as oxidative stress markers. Superoxide dismutase (SOD) and glutathione peroxidase (GPx) activities were measured as antioxidants. CIMT was assessed by carotid artery ultrasonography. 8‐OHdG/dG ratios and MDA levels were higher; SOD and GPx activities were lower in HD patients compared to controls. HD patients had significantly higher CIMT compared to controls (0.61 ± 0.08 vs. 0.42 ± 0.05, p < 0.001). There was a significant positive correlation between CIMT and 8‐OHdG/dG ratio (r = 0.57, p < 0.01) and MDA levels (r = 0.41, p < 0.01), while there was a significant negative correlation between CIMT and SOD (r = ?0.47, p < 0.01) and GPx levels (r = ?0.62, p < 0.01). It is firstly demonstrated that CIMT is positively correlated with oxidative DNA damage in HD patients without known atherosclerotic disease.     

基于局部约束邻域嵌入的人脸画像照片合成

针对画像块和照片块在流形上的邻域关系并不能完全反映彼此内在数据结构的问题,提出一种基于局部约束邻域嵌入(LCNE)的画像-照片合成算法。首先,利用基于邻域嵌入(NE)的合成方法得到待合成照片或画像的初始估计;其次,根据待合成的照片块或画像块与训练集中的照片块或画像块的相似性来约束合成权值;然后,通过交替优化方法进行权值的确定和K-近邻的选择,并更新待合成目标块;最后,合并所有估计的照片块或画像块合成目标图像。与基于邻域嵌入的画像照片合成方法相比,所提方法合成图像的结构相似度提高0.0503,脸识别准确率提高14%。实验结果表明,该方法解决了基于NE方法导致的邻域之间兼容性不强的问题,能大大减少合成图像上的噪声及变形。