Minimum Message Length in Hybrid ARMA and LSTM Model Forecasting

Modeling and analysis of time series are important in applications including economics, engineering, environmental science and social science. Selecting the best time series model with accurate parameters in forecasting is a challenging objective for scientists and academic researchers. Hybrid models combining neural networks and traditional Autoregressive Moving Average (ARMA) models are being used to improve the accuracy of modeling and forecasting time series. Most of the existing time series models are selected by information-theoretic approaches, such as AIC, BIC, and HQ. This paper revisits a model selection technique based on Minimum Message Length (MML) and investigates its use in hybrid time series analysis. MML is a Bayesian information-theoretic approach and has been used in selecting the best ARMA model. We utilize the long short-term memory (LSTM) approach to construct a hybrid ARMA-LSTM model and show that MML performs better than AIC, BIC, and HQ in selecting the model—both in the traditional ARMA models (without LSTM) and with hybrid ARMA-LSTM models. These results held on simulated data and both real-world datasets that we considered.We also develop a simple MML ARIMA model.     

Study of Nonlinear Models of Oscillatory Systems by Applying an Intelligent Computational Technique

In this paper, we have analyzed the mathematical model of various nonlinear oscillators arising in different fields of engineering. Further, approximate solutions for different variations in oscillators are studied by using feedforward neural networks (NNs) based on the backpropagated Levenberg–Marquardt algorithm (BLMA). A data set for different problem scenarios for the supervised learning of BLMA has been generated by the Runge–Kutta method of order 4 (RK-4) with the “NDSolve” package in Mathematica. The worth of the approximate solution by NN-BLMA is attained by employing the processing of testing, training, and validation of the reference data set. For each model, convergence analysis, error histograms, regression analysis, and curve fitting are considered to study the robustness and accuracy of the design scheme.     

Whether the Support Region of Three-Bit Uniform Quantizer Has a Strong Impact on Post-Training Quantization for MNIST Dataset?

Driven by the need for the compression of weights in neural networks (NNs), which is especially beneficial for edge devices with a constrained resource, and by the need to utilize the simplest possible quantization model, in this paper, we study the performance of three-bit post-training uniform quantization. The goal is to put various choices of the key parameter of the quantizer in question (support region threshold) in one place and provide a detailed overview of this choice’s impact on the performance of post-training quantization for the MNIST dataset. Specifically, we analyze whether it is possible to preserve the accuracy of the two NN models (MLP and CNN) to a great extent with the very simple three-bit uniform quantizer, regardless of the choice of the key parameter. Moreover, our goal is to answer the question of whether it is of the utmost importance in post-training three-bit uniform quantization, as it is in quantization, to determine the optimal support region threshold value of the quantizer to achieve some predefined accuracy of the quantized neural network (QNN). The results show that the choice of the support region threshold value of the three-bit uniform quantizer does not have such a strong impact on the accuracy of the QNNs, which is not the case with two-bit uniform post-training quantization, when applied in MLP for the same classification task. Accordingly, one can anticipate that due to this special property, the post-training quantization model in question can be greatly exploited.     

Passivation of aluminium–n+ silicon contacts for solar cells by ultrathin Al2O3 and SiO2 dielectric layers

Ultra‐thin thermally grown SiO₂ and atomic‐layer‐deposited (ALD) Al₂O₃ films are trialled as passivating dielectrics for metal–insulator–semiconductor (MIS) type contacts on top of phosphorus diffused regions applicable to high efficiency silicon solar cells. An investigation of the optimum insulator thickness in terms of contact recombination factor J_{0_cont} and contact resistivity ρ_c is undertaken on 85 Ω/□ and 103 Ω/□ diffusions. An optimum ALD Al₂O₃ thickness of ～22 Å produces a J_{0_cont} of ～300 fAcm^–2 whilst maintaining a ρ_c lower than 1 mΩ cm² for the 103 Ω/□ diffusion. This has the potential to improve the open‐circuit voltage by a maximum 15 mV. The thermally grown SiO₂ fails to achieve equivalently low J_{0_cont} values but exhibits greater thermal stability, resulting in slight improvements in ρ_c when annealed for 10 minutes at 300 °C without significant changes in J_{0_cont}. The after‐anneal J_{0_cont} reaches ～600 fAcm^–2 with a ρ_c of ～2.5 mΩ cm² for the 85 Ω/□ diffusion amounting to a maximum gain in open‐circuit voltage of 6 mV. (© 2013 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Neutron diffraction experiments in diamond and sapphire anvil cells

Diamond anvil cells (DAC) provide the highest static pressures ≥1?Mbar. Because of the low intensity of neutron sources, for a long time it was thought impossible to use DAC or other anvil cells in neutron experiments. We describe pressure cells with diamond and sapphire anvils and neutron instrumentation allowing neutron diffraction experiments to be carried out under pressures as high as 50?GPa, temperatures down to 0.1?K, and applied magnetic fields up to 7.5?T.     

Effect of gasketing and assembly design: a novel 10/3.5 mm multi-anvil assembly reaching perovskite pressures

A new design of multi-anvil assembly and modified gasket characteristics with octahedron and truncation edge lengths of 10 and 3.5 mm is presented for reaching pressures and temperatures over 24 GPa and 2000 °C, respectively. Partially dehydroxylated pyrophyllite half-gaskets with a tapered design fully nesting the octahedron have been employed to prevent excessive octahedron extrusion between the cubes. The assembly utilizes an axially placed thermocouple through the octahedral center, allowing two samples to be present at identical high P–T conditions on either side of the thermocouple during a run. A third sample can be used as a packing around the thermocouple, so long as that sample is inert with respect to the thermocouple and surrounding material. The temperature gradient within the sample locations has been well characterized using two-pyroxene thermometry in the CaO–MgO–SiO₂ system and numerical modeling calculations. The results indicate a good agreement in gradient shape, although the numerical model appears to under-estimate the magnitude of temperature change. The assembly maintains stable temperatures and provides low failure rates.     

Eine Dichtigkeitsprüfanlage für Halbleiterbauelemente auf der Grundlage von 85Kr

Auf der Basis von ⁸⁵Kr wurde ein Prüfstand für die routinemäßge Dichtigkeitsprüfung von Halbleiterbauelementen entwickelt und gebaut. Die Anlage arbeitet mit einer Aktivkohlepumpe und ist für einen Maximaldruck ion 13 kp/cm² und eine Maximalaktivität von 40Ci⁸⁵Kr ausgelegt. Es werden einige theoretische Angaben zur Methode gemacht und der Auf ban sowie die Funktionsweise des Prüfstandes beschrieben. Die erhaltenen Ergebnisse werden mit denen anderer Methoden verglichen.     

Zur quantitativen Auswertung von Mikro-Autoradiogrammen mit dem Densitron

Die vorliegende Arbeit beschāftigt sich mit der Answertung von Autoradiogrammen inhomogen markierter mikroskopischer Strukturen am Densitron. Dazu fertigten wir von herkömmlichen Stripping-Filmen (z. B. ORW O K 106) Mikro-Stufengraukeile an und bestimmten die Schwärzung S_ijeder Stufe photometrisch. Graukeile geeigneten Schwärzungsumfangs dienten dann zur Einstellung der Farbtrigger des Densitrons, so daß sich die bei dieser Einstellung gemessenen Farbflächen F_i des jeweiligen Objektes bestimmten Schwärzungswerten zuordnen ließen.

Als Modellobjekte verwendelen wir erstens Autoradiogramme von ³H-Thymidin-markierten Tumorzellen. Vergleichsmessungen an unterschiedlich lange exponierten, sonst aber identischen Präparaten ergaben eine gute Reproduzierbarkeit der Meßergebnisse. Als zweites Testobjekt dienlen Autoradiogramme einer nur schwach markierten Struktur im Gehirn von Fröschen. Die an diesem Grenzfall erhaltenen Ergebnisse werden mit Meßwerten verglichen, die vom gleichen Objekt mittels Scanning-Photometrie an einem. Mikroskop-Photometer SMP 01 (OPTON, Oberkoches, BRD) gewonnen wurden.     

介导GAD基因的人泡沫病毒载体的构建及其应用

构建了两种新的重组人泡沫病毒载体pGPSNI—hGAD67和pGPSNI—hGAD65，分别携带人符氨酸脱羧酶（GAD）的两种同工酶GAD65和GAD67基因，并研究其在帕金森病（PD）模刭鼠丘脑底核中的表达及对帕金森病的治疗作用．RT—PCR结果表明，人泡沫病毒载体成功地介导了GAD65基因和GAD67基因在PD模型鼠丘脑底核中有效表达．动物行为检测表明，GAD65基因导入组PD模型鼠行为与对照组相比得到明显的改善（n=12，由〈0．01）．