High-quality texture compression using adaptive color grouping and selection algorithm

In Quorum based cross datacenter distributed storage systems, data replication is a widely used technology. Such systems usually choose weak consistency, means they prone to select less number of replicas in read request in order to improve the performance, which lead to the possibility of getting stale data. In this paper, we propose an adaptive replica selection algorithm to make sure the system does not exceed specific stale read ratio while achieving good performance. Firstly, we choose a suitable distribution model from many candidates to estimate the time interval between current read request and nearest write request before it. Then we use Monte-Carlo method to simulate the processing order of read request and write request in different replicas. At last we use those estimated values to determine minimal number of replicas each read request needs to select in order to achieve a specific consistency level. We conduct a comprehensive experiments and the result shows our algorithm is effective.

Editor-in-Chief of Serbian Journal of Electrical Engineering1 requests to correct the first page of the paper A Multi-Layered Image Format for the Web with an Adaptive Layer Selection Algorithm by Milan Tair, Aleksandar Mihajlovic, Nikola Savanovic, Marko Sarac published in the Serbian journal of Electrical Engineering, Vol. 14, No. 2, June 2017, pp. 177 - 197, DOI: https://doi.org/10.2298/SJEE161010001T, since the text is missing in the footnote. The missing text is: An earlier version of this paper was presented at the International Scientific Conference on ICT AND E-Business Related Research, SINTEZA 2016, April 22, 2016 at Singidunum University in Belgrade, Serbia. <br><br><font color="red"><b> Link to the corrected article <u><a href="http://dx.doi.org/10.2298/SJEE161010001T">10.2298/SJEE161010001T</a></b></u>

In this paper, we present a scheme to improve immune cloning selection algorithm. The improved algorithm, which is referred to as adaptive cloning selection algorithm (ACSA), is proposed and then applied to function optimization. At first, we present adaptive gene mutation which decides mutation probability of each code point (locus) based on the quality of antibodies and the number of evolution iterations. Secondly, we present an iteratively increasing method from one locus to the all ones, which can be used in function optimization. Then, the cloning selection process of evolution is divided into two-stages. The first step is to increase locus of antibodies. In the other one, the Baldwin effect learning operator is employed. And finally, an experiment is carried out to verify the theoretical analyses on several testing functions.

The performance of underwater acoustic target classification decreases and is unstable when the training set contains noisy, redundant or irrelevant samples. In this paper, a new adaptive immune clonal sample selection algorithm (AICISA) is proposed to address this problem. AICISA is aimed at directing generation evolution. An experiment about the application of AICISA using the multi-field features extracted from 4 kinds of underwater acoustic targets was conducted. Experimental results show that AICISA can select effective subsets of samples. Reducing the sample size by 90%, the classification accuracy of SVM is improved by 10%. AICISA also shows good convergence and stability. The optimal subset of samples obtained by AICISA has good generalization ability and can remarkably reduce the classification time.

This paper presents a qualitative study in which we evaluate the core parts of an adaptive algorithm for next-exercise selection in an e-learning system. The algorithm was previously constructed from a series of studies where participants played the role of a teacher and chose the difficulty of a subsequent exercise for a learner based on their performance, mental effort and self esteem. In this paper, we present these findings to real teachers to gain insights into whether the algorithm is effective and appropriate for future inclusion in an intelligent tutoring system. Overall, we found that teachers believed that the recommendations from the algorithm were appropriate.

The main target of this article is to propose a simple Received Signal Strength Indicator‐ (RSSI‐) based adaptive Physical Layer (PHY) selection approach for Bluetooth Low Energy 5.0 (BLE 5) and compare the performance of adaptive PHY selection with the performance of individual PHY. This article also validates the simulation results acquired from the in‐house developed Ray Tracing (RT) tool with the BLE 5 measurements conducted in a corridor of the TUAS building, Aalto University. The performance metrics considered for the analysis are RSSI, throughput, and outage ratio. The impact of different transmission power on BLE 5 is also analyzed. The obtained simulation results indicate a Root Mean Square Error (RMSE) of about 4.6 dB between the measured and the simulated RSSI at 2.4 GHz frequency. By investigating these results, an adaptive PHY selection algorithm is proposed in this article. The adaptive PHY selection approach enhances the mean throughput of the system and improves the outage ratio as well. In the light of acquired results, the adaptive PHY selection scheme shows significant performance gain and is found more beneficial in the case of an environment with mixed RSSI samples, i.e., a mix of low and high RSSI values.

Fast analysis of molecular dynamics trajectories with graphics processing units—Radial distribution function histogramming

Cognitive radar is a new framework of radar system proposed by Simon Haykin recently. Adaptive waveform selection is an important problem of intelligent transmitter in cognit i ve radar. In this paper, the problem of adaptive waveform selection is modeled as stochastic dynamic programming model. Then backward dynamic programming, temporal difference learning and Q-learning are used to solve this problem. Optimal waveform selection algorithm and approximate solutions are proposed respectively. The simulation results demonstrate that the two approximate methods approach the optimal waveform selection scheme and have lower uncertainty of state estimation compared to fixed waveform. The performance of temporal difference learning is better than Q-learning, but Q-learning is more suitable to use in radar scene. Finally, the whole paper is summarized.

This paper proposes a joint adaptive rate selection and power control algorithm for broadband TDMA/TDM wireless networks. The proposed algorithm is a modified, enhanced, and more robust version of the selective power control with active link protection (SPC-ALP) algorithm proposed for adaptive transmission rate and power control in CDMA networks [R.Jantti and S. Kim, 2000]. Unlike SPC-ALP, the proposed algorithm uses adaptive coding and modulation (instead of variable spreading gain) for transmission rate control. In addition, the proposed algorithm is different in two aspects; the first one is the inclusion of the cochannel interferer assistance mode, while the second one is the use of the signal quality as a criterion for user removal. Results show that the proposed algorithm outperforms SPC-ALP in terms of the net throughput and signal quality measures such as the outage probability and frame error rate. Channel reallocation is also studied and found to be very effective in enhancing the system performance particularly at low to medium loading.

Polarization-sensitive array (PSA) can sense satellite signals from multiple dimensions and separate signals from interference according to various characteristics. Hence, the use of PSA improves the interference rejection capability of the global navigation satellite system (GNSS). Whereas it also brings a significant increase in computational complexity as there are more dimensions of array signal processing. Therefore, it is urgent to seek a method to employ fewer radiofrequency (RF) front-ends and antennas while still achieving superior anti-interference performance. In this paper, we investigate the influence of PSA configuration on interference suppression performance and propose an RF front-end optimal selection scheme for reconfigurable anti-jamming PSA with application to GNSS. Combining adaptive array processing algorithm and RF front-end selection strategy, we formulate the expression of the effective carrier to noise ratio (CNR) to analyze the performance enhancement generated by PSA reconfiguration. An approach is presented to solve the upper bound of the effective CNR to plot the trade-off curve between performance and cost, which can guide the selection of the optimal number of RF front-ends. Then another method is used to acquire the layout of the chosen optimal RF front-ends that can maximize the effective CNR. Numerical results illustrate the effectiveness and high efficiency of the proposed PSA reconfiguration scheme.

Extracting frequent itemsets from datasets is an important problem in data mining, for which several mining methods including FP‐Growth have been proposed. FP‐Growth is a classical frequent itemset mining method, which generates pattern databases without candidates. Many improvements have been made in the literature due to the high time complexity and memory usage of FP‐Growth. However, most of them still suffer from performance issues on large datasets. In this paper, we design an auxiliary structure, Array Prefix‐Tree (AP‐Tree), and propose a new algorithm, Array Prefix‐Tree Growth (APT‐Growth), which is further parallelized as a Spark workflow, referred to as PAPT‐Growth. Based on a density threshold, we incorporate an adaptive algorithm selection process into PAPT‐Growth to ensure its running time performance. We conduct extensive experiments on different thresholds and multiple datasets, and experimental results show the performance superiority of PAPT‐Growth in comparison with several state‐of‐the‐art methods such as PFP, YAFIM, and DFPS. The analysis on density reveals a changing point, which justifies the necessity and validity of adaptive algorithm selection.

Many tree algorithms have been developed for regression problems. Although they are regarded as good algorithms, most of them suffer from loss of prediction accuracy when there are many irrelevant variables and the number of predictors exceeds the number of observations. We propose the multistep regression tree with adaptive variable selection to handle this problem. The variable selection step and the fitting step comprise the multistep method. The multistep generalized unbiased interaction detection and estimation (GUIDE) with adaptive forward selection (fg) algorithm, as a variable selection tool, performs better than some of the well-known variable selection algorithms such as efficacy adaptive regression tube hunting (EARTH), FSR (false selection rate), LSCV (least squares cross-validation), and LASSO (least absolute shrinkage and selection operator) for the regression problem. The results based on simulation study show that fg outperforms other algorithms in terms of selection result and computation time. It generally selects the important variables correctly with relatively few irrelevant variables, which gives good prediction accuracy with less computation time.

Adaptive model selection for time series prediction in wireless sensor networks

Analyzing user behavior characteristics in a complex power grid environment is essential for user behavior planning and resource coordination optimization. Traditional user behavior analysis methods based on model-driven and causal analysis have the disadvantages of strong subjectivity and physical models that are difficult to deal with the randomness and uncertainty of user behavior in complex grid environments. In this paper, we use unsupervised learning methods to analyze user behavior in complex power grid environments, and propose user behavior analysis methods based on stacked autoencoder and clustering. We first reduce the complexity of user behavior data by proposing adaptive feature selection algorithm of user behavior based on stacked autoencoder and unsupervised learning (AFS-SAEUL). Finally, we build a user behavior analysis model based on adaptive feature selection and improved clustering (UBA-AFSIC). The model improved the performance of unsupervised classification of user behavior by fusing the adaptive generation strategy of the initial cluster centers. The simulation experiment results on two real electricity datasets and one public electric vehicle charging dataset show that compared with the existing feature selection algorithm and clustering algorithm, the algorithms proposed in this paper have higher feature selection rate and better clustering performance.

Adaptive operator scheduling algorithms for continuous query processing are usually designed to serve a single performance objective, such as minimizing memory usage or maximizing query throughput. We observe that different performance objectives may sometimes conflict with each other. Also due to the dynamic nature of streaming environments, the performance objective may need to change dynamically. Furthermore, the performance specification defined by users may itself be multi-dimensional. Therefore, utilizing a single scheduling algorithm optimized for a single objective is no longer sufficient. In this paper, we propose a novel adaptive scheduling algorithm selection framework named AMoS. It is able to leverage the strengths of existing scheduling algorithms to meet multiple performance objectives. AMoS employs a lightweight learning mechanism to assess the effectiveness of each algorithm. The learned knowledge can be used to select the algorithm that probabilistically has the best chance of improving the performance. In addition, AMoS has the flexibility to add and adapt to new scheduling algorithms, query plans and data sets during execution. Our experimental results show that AMoS significantly outperforms the existing scheduling algorithms with regard to satisfying both uni-objective and multi-objective performance requirements. 1