Search

This study investigated the nonvolatile memory characteristics of devices fabricated with PMMA embedding composite of graphene and molybdenum disulphide, which were employed as charge-trapping centres. X-ray diffraction measurements, Scanning and transmission electron microscope image analysis along with other characterisation techniques were performed for the assessment of the nature of graphene MoS2 hybrid system. Bistable I-V characteristics of the device revealed stable and rewritable memory effect. A significantly high resistance ratio > 104 (Roff/Ron) was observed even after 10 days. The reliability and reproducibility of the devices were tested for 104 cycles. The conduction mechanisms of the fabricated nanocomposite based memory cell were discussed on the basis of experimental data using a charge trapping–detrapping mechanism in the graphene MoS2 hybrid. Owing to the increasing interest in the flexible electronics, bending tests were carried out on the devices fabricated on PET substrates. Bendability test at various bending diameters (40-5 mm) for 100 cycles was carried out to show the mechanical robustness of the device.

Different weight percent (wt%) of Zinc Oxide (ZnO) nanorods, prepared by chemical route, are used into polymethylmethacrylate (PMMA) to fabricate organic bistable devices (OBDs). The devices utilize spin coated polymer nanocomopsite thin film, which exhibit electrical bistability and WORM characteristics in switching between high conductive (ON) and low conductive (OFF) states and it has been observed that along with ZnO nanorod concentration, oxygen vacancy content is also an important factor for the achievement of ON-OFF conductive states. The maximum ON/OFF current ratio is of the order of 5×103 for Al/ZnO nanorods:PMMA/ITO heterostructures with 0.5wt% of ZnO nanorods. Statistical results indicate that the operating stability of devices with high retention time of 103s. We demonstrate that the enhancement of memory parameters are controlled by the oxygen vacancies existed in the engineered ZnO nanorods. These native defects are credited as basis of resistive switching behaviour of the devices. By comparing the C-V curves to the current-voltage (I-V) curves, dielectric constant, electron mobility and oxide trap densities have been estimated. Display Omitted The device involved ZnO nanorods as active nanomaterial.The good optical-transparency, significant I-V and C-V hysteresis windows suggest a see through memory device.The oxygen vacancy engineered ZnO nanorods have been credited for the switching mechanism.The findings suggest a simple fabrication route towards the development of an optically transparent organic memory device.

Interfacial composition and electrical properties of ultra-thin hafnium oxide (HfO 2 ) films on p-type Si substrate are reported. Hafnium oxide (HfO 2 ) thin films are prepared using radio-frequency sputtering method and subsequently annealed at different temperature. The effect of post-deposition annealing on the interfacial and chemical state of HfO 2 /Si gate stack has been characterised by means of X-ray reflectivity and X-ray photoelectron spectroscopy studies. Peaks of X-ray photoelectron spectroscopy spectra at 530.50 and 532.25 eV originate from Hf–O–Si bond illustrated the creation of Hf-silicate based interfacial layer at the high-k/Si interface. X-ray reflectivity fitting result also corroborated the formation of Hf silicate interfacial layer. Capacitance-voltage measurements revealed insignificant hysteresis in case of film annealed at 600 °C. Interface trap density has been extracted using Terman method and is found to be 3.18×10 −11 cm −2 eV −1 at −1.0 V. Minimum equivalent oxide thickness (EOT) of 1.3 nm was obtained for the film annealed at 600° C. The gate leakage current density of the HfO 2 film annealed at 600 °C is 1.5×10 −5 A/cm 2 at a bias voltage of −2 V.

By controlling the voltage-polarity, a bipolar resistive switching (RS) behavior was observed in CVD grown tungsten disulfide (WS2) thin film memory device fabricated on Pt/Ti/SiO2/Si substrate with top Al contacts. The RS performance of the device did not degrade even after repetitive voltage cycles, which indicate uniformity and reproducibility. A decent on/off ratio, high retention time, and a large endurance were respectively recorded to establish the non-volatile memory characteristics in a device with thin layer of WS2. The XRD, Raman, FESEM and EDAX measurement were included as the structural and compositional details of CVD grown WS2 thin film. The current conduction mechanisms were discussed on the basis of experimental data using various existing fitting models. This work also explained the mechanism of resistance switching in CVD grown Tungsten Disulfide. The findings offered a feasible approach towards fabrication and application of the Al/WS2/Pt/Ti/SiO2/Si device.

Hexagonal zinc oxide nanotetrapods structures were grown on p-type Si substrates by the simple vapour-solid technique without the presence of any catalysts. Field emission scanning electron microscopy and X-ray diffraction techniques were used to study the morphologies and crystal structures of these tetrapod-shaped ZnO nanocrystals. The photoluminescence measurements of ZnO tetrapods exhibited UV peak at 380 nm corresponding to the free exciton emission, and a broad peak at about 506 nm arising out of defect related emission from surface of the nanocrystals. Furthermore, temperature dependent photoluminescence characteristics of ZnO tetrapods have been investigated and origin of LO phonon emission is explained in terms of both bound and free excitons. Current–voltage (I–V) characteristics of the ZnO tetrapods devices were analysed, and they exhibited growth-temperature dependent electrical property. Depending on the density of grown nanostructures the I–V of the ZnO/p-Si heterostructure was varied from Ohmic to rectifying behaviour and current transport mechanism is found to follow space charge limited current.

With the advent of high-dimensional stored big data and streaming data, suddenly machine learning on a very large scale has become a critical need. Such machine learning should be extremely fast, should scale up easily with volume and dimension, should be able to learn from streaming data, should automatically perform dimension reduction for high-dimensional data, and should be deployable on hardware. Neural networks are well positioned to address these challenges of large scale machine learning. In this paper, we present a method that can effectively handle large scale, high-dimensional data. It is an online method that can be used for both streaming and large volumes of stored big data. It primarily uses Kohonen nets, although only a few selected neurons (nodes) from multiple Kohonen nets are actually retained in the end; we discard all Kohonen nets after training. We use Kohonen nets both for dimensionality reduction through feature selection and for building an ensemble of classifiers using single Kohonen neurons. The method is meant to exploit massive parallelism and should be easily deployable on hardware that implements Kohonen nets. Some initial computational results are presented.

This paper presents methods for training pattern (prototype) selection, class-specific feature selection and classification for automated learning. For training pattern selection, we propose a method of sampling that extracts a small number of representative training patterns (prototypes) from the dataset. The idea is to extract a set of prototype training patterns that represents each class region in a classification problem. In class-specific feature selection, we try to find a separate feature set for each class such that it is the best one to separate that class from the other classes. We then build a separate classifier for that class based on its own feature set. The paper also presents a new hypersphere classification algorithm. Hypersphere nets are similar to radial basis function (RBF) nets and belong to the group of kernel function nets. Polynomial time complexity of the methods is proven. Polynomial time complexity of learning algorithms is important to the field of neural networks. Computational results are provided for a number of well-known datasets. None of the parameters of the algorithm were fine tuned for any of the problems solved and this supports the idea of automation of learning methods. Automation of learning is crucial to wider deployment of learning technologies.

This paper presents a new algorithm for generating radial basis function (RBF)-like nets for classification problems. The method uses linear programming (LP) models to train the RBF-like net. Polynomial time complexity of the method is proven and computational results are provided for many well-known problems. The method can also be implemented as an on-line adaptive algorithm.

This paper presents a polynomial time algorithm for the construction and training of a class of multilayer perceptrons for classification. It uses linear programming models to incrementally generate the hidden layer in a restricted higher-order perceptron. Polynomial time complexity of the method is proven. Computational results are provided for several well-known applications in the areas of speech recognition, medical diagnosis, and target detection. In all cases, very small nets were created that had error rates similar to those reported so far.

The author of IFPS/OPTIMUM presents some typical applications of optimization within decision support systems (DSS). They show that finding ‘what's best’ is gradually superseding the ‘what if’ and ‘goal seeking’ type of analysis that is prevalent now. IFPS/OPTIMUM is now used by over 130 companies, including many of the Fortune 500 ones. The main drawback of ‘what if’ and ‘goal seeking’ type of analysis is that one cannot impose restrictions on model variables and have the DSS generator solve for values of the ‘what if’ or ‘goal seek’ variables that satisfy those restrictions. Instead, one would have to muddle through many ‘what if's’ or ‘goal seeks’ at random and check if the restrictions are satisfied. It is a manual process that is very time consuming. ‘What's best’ analysis with optimization allows one to impose such restrictions easily and do the analysis quickly. This type of analysis is playing an increasingly significant role within decision support systems. Practicing management scientists and managers should note this phenomenon. New tools and new ways of addressing problems are emerging.

Automatic classification of optimization problems as linear or nonlinear is now being demanded in some modeling and optimization systems to make such systems more user friendly. This requires that the system examine the model and make such determination. On such method of analysis is discussed in this paper.

End-user optimization is becoming more widespread. Current software coupling optimization with spreadsheet systems now makes it possible for a user without a mathematical background to develop useful and substantive decision support tools utilizing mathematical programming. This paper briefly compares three existing spreadsheet optimization systems, and describes industrial and educational experience with one, ifps /Optimum. Problems encountered in using such systems are also described.

The paper disproves the general belief that one should not acquire companies with a higher PE ratio than one's own under a stock exchange, since it would dilute the acquirer's earnings per share. It develops new models for calculating the earnings per share and share value for acquirer's shareholders allowing for partial acquisitions and synergistic effects. It then determines the maximum PE multiple that the acquirer can pay that will neither dilute the share value nor the projected earnings per share. It also establishes that the maximum PE multiple one can pay in either case is highest when only a majority of shares is acquired. The condition for the maximum PE multiple to be greater than the acquirer's PE ratio is also derived. A bargaining range for the PE multiple is determined too.