Machine Learning with Dimensionality Reduction for DDoS Attack Detection

With the advancement of internet, there is also a rise in cybercrimes and digital attacks. DDoS (Distributed Denial of Service) attack is the most dominant weapon to breach the vulnerabilities of internet and pose a significant threat in the digital environment. These cyber-attacks are generated deliberately and consciously by the hacker to overwhelm the target with heavy traffic that genuine users are unable to use the target resources. As a result, targeted services are inaccessible by the legitimate user. To prevent these attacks, researchers are making use of advanced Machine Learning classifiers which can accurately detect the DDoS attacks. However, the challenge in using these techniques is the limitations on capacity for the volume of data and the required processing time. In this research work, we propose the framework of reducing the dimensions of the data by selecting the most important features which contribute to the predictive accuracy. We show that the ‘lite’ model trained on reduced dataset not only saves the computational power, but also improves the predictive performance. We show that dimensionality reduction can improve both effectiveness (recall) and efficiency (precision) of the model as compared to the model trained on ‘full’ dataset.     

In Vitro Selection of New DNA Aptamers for Human Vascular Endothelial Growth Factor 165

Two DNA aptamers that bind the heparin-binding domain (HBD) of the human vascular endothelial growth factor 165 (VEGF-165) have been previously reported. Although VEGF-165 is a homodimeric protein and the two aptamers have different sequences and secondary structures, the aptamers appear to occupy the same binding site and cannot form a 2 : 1 aptamer/protein complex, thus making them unsuitable for creating a higher-affinity dimeric DNA aptamer. This has motivated us to conduct a new in vitro selection experiment to search for new VEGF-165-binding DNA aptamers with different properties. We undertook a multistream selection strategy in which the concentration of VEGF-165 was varied significantly. We carried out 11 rounds of selection, and next-generation sequencing was conducted for every round in each stream. From comprehensive sequence analysis, we identified four classes of DNA aptamers, of which two were reported before, but two are new DNA aptamers. One of the new aptamers exhibits a unique property that has never been observed before: it is capable of forming the 2 : 1 aptamer/protein complex with VEGF-165. This work has expanded the repertoire of VEGF-165-binding DNA aptamers and creates a possibility to engineer a higher affinity homodimeric aptamer for VEGF-165.     

Analyse von Tilgerkonfigurationen mit trockenen Reibungselementen

Forschung im Ingenieurwesen - Die Ergreifung von Maßnahmen zur Schwingungsreduktion ist eine im Maschinenbau wichtige Aufgabe, da unerwünschte Vibrationen zu Systemstörungen...     

Liquid pump-enabled hydrogen refueling system for heavy duty fuel cell vehicles: Pump performance and J2601-compliant fills with precooling

We have developed a hydrogen (H₂) refueling solution capable of delivering precooled, compressed gaseous hydrogen for heavy duty vehicle (HDV) refueling applications. The system uses a submerged pump to deliver pressurized liquid H₂ from a cryogenic storage tank to a dispensing control loop that vaporizes the liquid and adjusts the pressure and temperature of the resulting gas to enable refueling at 35 MPa and temperatures as low as ?40 °C. A full-scale mobile refueler was fabricated and tested over a 6-month campaign to validate its performance. We report results from tests involving a total of 9000 kg of liquid H₂ pumped and 1350 filling cycles over a range of conditions. Notably, the system was able to repeatably complete multiple, back-to-back 30 kg filling cycles in under 6 min each, in full compliance with the SAE J2601-2 standard, demonstrating its potential for rapid-throughput HDV refueling applications.     

Preparation and characterization of PS‐PMMA/ZnO nanocomposite films with novel properties of high transparency and UV‐shielding capacity

High transparent and UV‐shielding poly (styrene)‐co‐poly(methyl methacrylate) (PS‐PMMA)/zinc oxide (ZnO) optical nanocomposite films were prepared by solution mixing using methyl ethyl ketone (MEK) as a cosolvent. The films were characterized by X‐ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), ultraviolet–visible (UV–vis) spectra, high‐resolution transmission electron microscopy (HR‐TEM), and atomic force microscope (AFM). Cross‐section HR‐TEM and AFM images showed that the ZnO nanoparticles were uniformly dispersed in the polymer matrix at the nanoscale level. The XRD and FTIR studies indicate that there is no chemical bond or interaction between PS‐PMMA and ZnO nanoparticles in the nanocomposite films. The UV–vis spectra in the wavelength range of 200–800 nm showed that nanocomposite films with ZnO particle contents from 1 to 20 wt % had strong absorption in UV spectrum region and the same transparency as pure PMMA‐PS film in the visible region. The optical properties of polymer are greatly improved by the incorporation of ZnO nanoparticles. © 2010 Wiley Periodicals, Inc. J Appl Polym Sci, 2010     

Evaluation of metal-nanowire electrical contacts by measuring contact end resistance

It is known, but often unappreciated, that the performance of nanowire (NW)-based electrical devices can be significantly affected by electrical contacts between electrodes and NWs, sometimes to the extent that it is really the contacts that determine the performance. To correctly understand and design NW device operation, it is thus important to carefully measure the contact resistance and evaluate the contact parameters, specific contact resistance and transfer length. A four-terminal pattern or a transmission line model (TLM) pattern has been widely used to measure contact resistance of NW devices and the TLM has been typically used to extract contact parameters of NW devices. However, the conventional method assumes that the electrical properties of semiconducting NW regions covered by a metal are not changed after electrode formation. In this study, we report that the conventional methods for contact evaluation can give rise to considerable errors because of an altered property of the NW under the electrodes. We demonstrate that more correct contact resistance can be measured from the TLM pattern rather than the four-terminal pattern and correct contact parameters including the effects of changed NW properties under electrodes can be evaluated by using the contact end resistance measurement method.     

Stress softening of multigraft copolymers

The hysteresis behaviour of multigraft (MG) copolymers, with a polyisoprene backbone and polystyrene (PS) side chains, was investigated by applying a modified softening model proposed by Elías-Zúñiga, which uses an approach of Ogden and Roxburgh. The model was combined with the non-affine tube model of rubber elasticity of Kaliske and Heinrich. Four parameters are obtained: chemical and physical cross-link moduli (G_c, G_e), the number of statistical segments between two successive entanglements (n_e/T_e) and a softening parameter (b). The model was proven to be valid by a comparison with other methods evaluating hysteresis behaviour. The characterization of the multigraft copolymers revealed a branch point and molecular architecture dependence of the softening parameter. b was low for tetrafunctional MG copolymers with cylindrical microdomains, and it was further reduced for a spherical morphology and for more complex molecular architectures. The magnitude of b also depends on the PS arm molecular weight for hexa- and tetrafunctional multigraft copolymers.