Energy efficient fuzzy-based DASH adaptation algorithm

With the expansive demand for video streaming over mobile networks, it is necessary to adopt schemes that balance the need for high video quality with the available network resources when streaming or downloading the video. Several approaches were proposed in the literature, including Dynamic Adaptive Streaming over HTTP (DASH). In this work, we consider an approach in which we place sufficient emphasis on the constrained battery resources in mobile devices when making decisions on the quality (or bitrate) of the video to be requested. This is done by using a fuzzy logic controller that enhances the performance of the Fuzzy-based DASH (FDASH) scheme. Simulation results show that our proposed approach conserves more energy than its predecessor while maintaining similar video quality and avoiding playback interruptions.     

A Framework for the Verification of Certifying Computations

Formal verification of complex algorithms is challenging. Verifying their implementations goes beyond the state of the art of current automatic verification tools and usually involves intricate mathematical theorems. Certifying algorithms compute in addition to each output a witness certifying that the output is correct. A checker for such a witness is usually much simpler than the original algorithm—yet it is all the user has to trust. The verification of checkers is feasible with current tools and leads to computations that can be completely trusted. We describe a framework to seamlessly verify certifying computations. We use the automatic verifier VCC for establishing the correctness of the checker and the interactive theorem prover Isabelle/HOL for high-level mathematical properties of algorithms. We demonstrate the effectiveness of our approach by presenting the verification of typical examples of the industrial-level and widespread algorithmic library LEDA.     

The Cognitive Benefit of Dynamic Representations on Procedural Skill Acquisition: A Computational Modeling Approach

Cognitive computational modeling is a viable methodology for further investigation of the hitherto inconclusive findings on the cognitive benefits of dynamic versus static visualization components of instructions. This is more so as contemporary cognitive architectures such as the Adaptive Control of Thought–Rational (ACT–R) 6.0 are increasingly applied to traditional cognitive psychology research problems. The application of this methodology is, however, restricted by the limited capability of existing architectures for implementing detailed atomic motor actions such as those involved in complex skill acquisition and performance. This article presents a 2-component computational modeling methodology for investigating the cognitive processes involved in the acquisition and performance of skilled motor tasks. The approach specifies a novel combination of a sequence-of-point technique with a movement control mechanism to implement variously acquired cognitive mental task representations and their intertwined role in postlearning performance as evident in the atomic control of motor actions. This paradigm is validated for 2 experiments using incrementally developed cognitive models developed in ACT–R 6.0. The model's quantitative outputs correlate significantly with equivalent empirical human data. This has implications for multimedia instructional design, especially where rapid, transferrable skill acquisition is desired on initial exposure.     

A fine-grained Random Forests using class decomposition: an application to medical diagnosis

Class decomposition describes the process of segmenting each class into a number of homogeneous subclasses. This can be naturally achieved through clustering. Utilising class decomposition can provide a number of benefits to supervised learning, especially ensembles. It can be a computationally efficient way to provide a linearly separable data set without the need for feature engineering required by techniques like support vector machines and deep learning. For ensembles, the decomposition is a natural way to increase diversity, a key factor for the success of ensemble classifiers. In this paper, we propose to adopt class decomposition to the state-of-the-art ensemble learning Random Forests. Medical data for patient diagnosis may greatly benefit from this technique, as the same disease can have a diverse of symptoms. We have experimentally validated our proposed method on a number of data sets that are mainly related to the medical domain. Results reported in this paper show clearly that our method has significantly improved the accuracy of Random Forests.     

Deep imitation learning for 3D navigation tasks

Deep learning techniques have shown success in learning from raw high-dimensional data in various applications. While deep reinforcement learning is recently gaining popularity as a method to train intelligent agents, utilizing deep learning in imitation learning has been scarcely explored. Imitation learning can be an efficient method to teach intelligent agents by providing a set of demonstrations to learn from. However, generalizing to situations that are not represented in the demonstrations can be challenging, especially in 3D environments. In this paper, we propose a deep imitation learning method to learn navigation tasks from demonstrations in a 3D environment. The supervised policy is refined using active learning in order to generalize to unseen situations. This approach is compared to two popular deep reinforcement learning techniques: deep-Q-networks and Asynchronous actor-critic (A3C). The proposed method as well as the reinforcement learning methods employ deep convolutional neural networks and learn directly from raw visual input. Methods for combining learning from demonstrations and experience are also investigated. This combination aims to join the generalization ability of learning by experience with the efficiency of learning by imitation. The proposed methods are evaluated on 4 navigation tasks in a 3D simulated environment. Navigation tasks are a typical problem that is relevant to many real applications. They pose the challenge of requiring demonstrations of long trajectories to reach the target and only providing delayed rewards (usually terminal) to the agent. The experiments show that the proposed method can successfully learn navigation tasks from raw visual input while learning from experience methods fail to learn an effective policy. Moreover, it is shown that active learning can significantly improve the performance of the initially learned policy using a small number of active samples.

     

Delivery of Nucleic Acids through the Controlled Disassembly of Multifunctional Nanocomplexes

In this study, novel pH‐responsive polyion complex micelles (PICMs) were developed for the efficient delivery of nucleic acid drugs, such as antisense oligonucleotide (AON) and short interfering RNA (siRNA). The PICMs consisted of a poly(amidoamine) (PAMAM) dendrimer–nucleic acid core and a detachable poly(ethylene glycol)‐block‐poly( propyl methacrylate‐co‐methacrylic acid) (PEG‐b‐P(PrMA‐co‐MAA)) shell. The micelles displayed a mean hydrodynamic diameter ranging from 50 to 70 nm, a narrow size distribution, and a nearly neutral surface charge. They could be lyophilized without any additives and stored in dried form. Upon redispersion in water, no change in complexation efficiency or colloidal properties was observed. Entry of the micelles into cancers cells was mediated by a monoclonal antibody fragment positioned at the extremity of the PEG segment via a disulfide linkage. Upon cellular uptake and protonation of the MAA units in the acidic endosomal environment, the micelles lost their corona, thereby exposing their positively charged endosomolytic PAMAM/nucleic acid core. When these pH‐responsive targeted PICMs were loaded with AON or siRNAs that targeted the oncoprotein Bcl‐2, they exhibited a greater transfection activity than nontargeted PICMs or commercial PAMAM dendrimers. Moreover, their nonspecific cytotoxicity was lower than that of PAMAM. The pH‐responsive PICMs reported here appear as promising carriers for the delivery of nucleic acids.     

Concentrator silicon solar cells from silicon industry rejects

Two types of silicon (Si) substrates (40 n-type with uniform base doping and 40 n/n⁺ epitaxial wafers) from the silicon industry rejects were chosen as the starting material for low-cost concentrator solar cells. They were divided into four groups, each consisting of 20 substrates: 10 are n/n⁺ and 10 are n substrates, and the solar cells were prepared for different diffusion times (45, 60, 75 and 90 min). The fabricated solar cells on n/n⁺ substrates (prepared with a diffusion time of 75 min) showed better parameters. In order to improve their performances, particularly the fill factor, 20 new solar cells on n/n⁺ substrates were fabricated using the same procedure (the diffusion time was 75 min)—but with four new front contact patterns. Investigation of current–voltage (I–V) characteristics under AM 1.5 showed that the parameters of these 20 new solar cells have improved in comparison to previous solar cells' parameters, and were as follows: open-circuit voltage (V_OC=0.57 V); short circuit current (I_SC=910 mA), and efficiency (η=9.1%). Their fill factor has increased about 33%. The I–V characteristics of these solar cells were also investigated under different concentration ratios (X), and they exhibited the following parameters (under X=100 suns): V_OC=0.62 V and I_SC=36 A.     

Wind resource assessment of the Jordanian southern region

Wind data in terms of annual, seasonal and diurnal variations at Queira, which is located in the southern part of Jordan was studied and analyzed. For this purpose, long-term wind speed data for a period of 12 years (1990–2001) was used. The analysis showed that the seasonal and diurnal pattern of wind speed matches the electricity load pattern of the location. Higher winds of the order of 6 m/s and more were observed during both the summer months of the year (May–August) and peak hours (1100–1500) of the day. The wind duration availability is discussed as the number of hours during which the wind remained in certain wind speed intervals. The possibility of electricity generation from wind power at Queira was carried out using three different wind energy systems of sizes 100, 22 kW rated power, and a wind farm consisting of 25 small wind turbines; each of 4 kW rated power with hub heights of 20, 30, and 40 m. The energy production analysis showed higher production from the wind farm with a 20 m hub height than the production from the other two wind turbines. Similarly, the cost analysis showed that the lowest generation costs of 1 kWh were obtained for the wind farm compared to the other two wind turbines. The possibility of water pumping using the wind farm was also investigated. The results showed that water pumping using wind turbines is an appropriate alternative for the photovoltaic water pumping in the region.