Learning from time series:Supervised Aggregative Feature Extraction

Many modeling problems require to estimate a scalar output from one or more time series. Such problems are usually tackled by extracting a fixed number of features from the time series (like their statistical moments), with a consequent loss in information that leads to suboptimal predictive models. Moreover, feature extraction techniques usually make assumptions that are not met by real world settings (e.g. uniformly sampled time series of constant length), and fail to deliver a thorough methodology to deal with noisy data. In this paper a methodology based on functional learning is proposed to overcome the aforementioned problems; the proposed Supervised Aggregative Feature Extraction (SAFE) approach allows to derive continuous, smooth estimates of time series data (yielding aggregate local information), while simultaneously estimating a continuous shape function yielding optimal predictions. The SAFE paradigm enjoys several properties like closed form solution, incorporation of first and second order derivative information into the regressor matrix, interpretability of the generated functional predictor and the possibility to exploit Reproducing Kernel Hilbert Spaces setting to yield nonlinear predictive models. Simulation studies are provided to highlight the strengths of the new methodology w.r.t. standard unsupervised feature selection approaches. © 2012 IEEE.

Efficient Marginal Likelihood Computation for Gaussian Process Regression

In a Bayesian learning setting, the posterior distribution of a predictive model arises from a trade-off between its prior distribution and the conditional likelihood of observed data. Such distribution functions usually rely on additional hyperparameters which need to be tuned in order to achieve optimum predictive performance; this operation can be efficiently performed in an Empirical Bayes fashion by maximizing the posterior marginal likelihood of the observed data. Since the score function of this optimization problem is in general characterized by the presence of local optima, it is necessary to resort to global optimization strategies, which require a large number of function evaluations. Given that the evaluation is usually computationally intensive and badly scaled with respect to the dataset size, the maximum number of observations that can be treated simultaneously is quite limited. In this paper, we consider the case of hyperparameter tuning in Gaussian process regression. A straightforward implementation of the posterior log-likelihood for this model requires O(N^3) operations for every iteration of the optimization procedure, where N is the number of examples in the input dataset. We derive a novel set of identities that allow, after an initial overhead of O(N^3), the evaluation of the score function, as well as the Jacobian and Hessian matrices, in O(N) operations. We prove how the proposed identities, that follow from the eigendecomposition of the kernel matrix, yield a reduction of several orders of magnitude in the computation time for the hyperparameter optimization problem. Notably, the proposed solution provides computational advantages even with respect to state of the art approximations that rely on sparse kernel matrices.

Process simulation of co-firing torrefied biomass in a 220 MWe coal-fired power plant

Torrefaction based co-firing in a pulverized coal boiler has been proposed for large percentage of biomass co-firing. A 220 MWe pulverized coal-power plant is simulated using Aspen Plus for full understanding the impacts of an additional torrefaction unit on the efficiency of the whole power plant, the studied process includes biomass drying, biomass torrefaction, mill systems, biomass/coal devolatilization and combustion, heat exchanges and power generation. Palm kernel shells (PKS) were torrefied at same residence time but 4 different temperatures, to prepare 4 torrefied biomasses with different degrees of torrefaction. During biomass torrefaction processes, the mass loss properties and released gaseous components have been studied. In addition, process simulations at varying torrefaction degrees and biomass co-firing ratios have been carried out to understand the properties of CO₂ emission and electricity efficiency in the studied torrefaction based co-firing power plant. According to the experimental results, the mole fractions of CO ₂ and CO account for 69-91% and 4-27% in torrefied gases. The predicted results also showed that the electrical efficiency reduced when increasing either torrefaction temperature or substitution ratio of biomass. A deep torrefaction may not be recommended, because the power saved from biomass grinding is less than the heat consumed by the extra torrefaction process, depending on the heat sources. 

Enhancing design space exploration by extending CPU/GPU specifications onto FPGAs

The design cycle for complex special-purpose computing systems is extremely costly and time-consuming. It involves a multiparametric design space exploration for optimization, followed by design verification. Designers of special purpose VLSI implementations often need to explore parameters, such as optimal bitwidth and data representation, through time-consuming Monte Carlo simulations. A prominent example of this simulation-based exploration process is the design of decoders for error correcting systems, such as the Low-Density Parity-Check (LDPC) codes adopted by modern communication standards, which involves thousands of Monte Carlo runs for each design point. Currently, high-performance computing offers a wide set of acceleration options that range from multicore CPUs to Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs). The exploitation of diverse target architectures is typically associated with developing multiple code versions, often using distinct programming paradigms. In this context, we evaluate the concept of retargeting a single OpenCL program to multiple platforms, thereby significantly reducing design time. A single OpenCL-based parallel kernel is used without modifications or code tuning on multicore CPUs, GPUs, and FPGAs. We use SOpenCL (Silicon to OpenCL), a tool that automatically converts OpenCL kernels to RTL in order to introduce FPGAs as a potential platform to efficiently execute simulations coded in OpenCL. We use LDPC decoding simulations as a case study. Experimental results were obtained by testing a variety of regular and irregular LDPC codes that range from short/medium (e.g., 8,000 bit) to long length (e.g., 64,800 bit) DVB-S2 codes. We observe that, depending on the design parameters to be simulated, on the dimension and phase of the design, the GPU or FPGA may suit different purposes more conveniently, thus providing different acceleration factors over conventional multicore CPUs.

Performance and Fault Tolerance of Preconditioned Iterative Solvers on Low-Power ARM Architectures

As the complexity of computing systems grows, reliability and energy are two crucial challenges asking for holistic solutions. In this paper, we investigate the interplay among concurrency, power dissipation, energy consumption and voltage-frequency scaling for a key numerical kernel for the solution of sparse linear systems. Concretely, we leverage a task-parallel implementation of the Conjugate Gradient method, equipped with an state-of-the-art pre-conditioner embedded in the ILUPACK software, and target a low-power multi core processor from ARM.In addition, we perform a theoretical analysis on the impact of a technique like Near Threshold Voltage Computing (NTVC) from the points of view of increased hardware concurrency and error rate.

ALEA: Fine-Grain Energy Profiling with Basic Block Sampling

Energy efficiency is an essential requirement for all contemporary computing systems. We thus need tools to measure the energy consumption of computing systems and to understand how workloads affect it. Significant recent research effort has targeted direct power measurements on production computing systems using on-board sensors or external instruments. These direct methods have in turn guided studies of software techniques to reduce energy consumption via workload allocation and scaling. Unfortunately, direct energy measurements are hampered by the low power sampling frequency of power sensors. The coarse granularity of power sensing limits our understanding of how power is allocated in systems and our ability to optimize energy efficiency via workload allocation.
We present ALEA, a tool to measure power and energy consumption at the granularity of basic blocks, using a probabilistic approach. ALEA provides fine-grained energy profiling via sta- tistical sampling, which overcomes the limitations of power sens- ing instruments. Compared to state-of-the-art energy measurement tools, ALEA provides finer granularity without sacrificing accuracy. ALEA achieves low overhead energy measurements with mean error rates between 1.4% and 3.5% in 14 sequential and paral- lel benchmarks tested on both Intel and ARM platforms. The sampling method caps execution time overhead at approximately 1%. ALEA is thus suitable for online energy monitoring and optimization. Finally, ALEA is a user-space tool with a portable, machine-independent sampling method. We demonstrate two use cases of ALEA, where we reduce the energy consumption of a k-means computational kernel by 37% and an ocean modelling code by 33%, compared to high-performance execution baselines, by varying the power optimization strategy between basic blocks.

Supervised Aggregative Feature Extraction for Big Data Time Series Regression

In many applications, and especially those where batch processes are involved, a target scalar output of interest is often dependent on one or more time series of data. With the exponential growth in data logging in modern industries such time series are increasingly available for statistical modeling in soft sensing applications. In order to exploit time series data for predictive modelling, it is necessary to summarise the information they contain as a set of features to use as model regressors. Typically this is done in an unsupervised fashion using simple techniques such as computing statistical moments, principal components or wavelet decompositions, often leading to significant information loss and hence suboptimal predictive models. In this paper, a functional learning paradigm is exploited in a supervised fashion to derive continuous, smooth estimates of time series data (yielding aggregated local information), while simultaneously estimating a continuous shape function yielding optimal predictions. The proposed Supervised Aggregative Feature Extraction (SAFE) methodology can be extended to support nonlinear predictive models by embedding the functional learning framework in a Reproducing Kernel Hilbert Spaces setting. SAFE has a number of attractive features including closed form solution and the ability to explicitly incorporate first and second order derivative information. Using simulation studies and a practical semiconductor manufacturing case study we highlight the strengths of the new methodology with respect to standard unsupervised feature extraction approaches.

A Fast Algorithm for Sparse Support Vector Machines for Mobile Computing Applications

This research presents a fast algorithm for projected support vector machines (PSVM) by selecting a basis vector set (BVS) for the kernel-induced feature space, the training points are projected onto the subspace spanned by the selected BVS. A standard linear support vector machine (SVM) is then produced in the subspace with the projected training points. As the dimension of the subspace is determined by the size of the selected basis vector set, the size of the produced SVM expansion can be specified. A two-stage algorithm is derived which selects and refines the basis vector set achieving a locally optimal model. The model expansion coefficients and bias are updated recursively for increase and decrease in the basis set and support vector set. The condition for a point to be classed as outside the current basis vector and selected as a new basis vector is derived and embedded in the recursive procedure. This guarantees the linear independence of the produced basis set. The proposed algorithm is tested and compared with an existing sparse primal SVM (SpSVM) and a standard SVM (LibSVM) on seven public benchmark classification problems. Our new algorithm is designed for use in the application area of human activity recognition using smart devices and embedded sensors where their sometimes limited memory and processing resources must be exploited to the full and the more robust and accurate the classification the more satisfied the user. Experimental results demonstrate the effectiveness and efficiency of the proposed algorithm. This work builds upon a previously published algorithm specifically created for activity recognition within mobile applications for the EU Haptimap project [1]. The algorithms detailed in this paper are more memory and resource efficient making them suitable for use with bigger data sets and more easily trained SVMs.

Acceleration-as-a-Service: Exploiting Virtualised GPUs for a Financial Application

How can GPU acceleration be obtained as a service in a cluster? This question has become increasingly significant due to the inefficiency of installing GPUs on all nodes of a cluster. The research reported in this paper is motivated to address the above question by employing rCUDA (remote CUDA), a framework that facilitates Acceleration-as-a-Service (AaaS), such that the nodes of a cluster can request the acceleration of a set of remote GPUs on demand. The rCUDA framework exploits virtualisation and ensures that multiple nodes can share the same GPU. In this paper we test the feasibility of the rCUDA framework on a real-world application employed in the financial risk industry that can benefit from AaaS in the production setting. The results confirm the feasibility of rCUDA and highlight that rCUDA achieves similar performance compared to CUDA, provides consistent results, and more importantly, allows for a single application to benefit from all the GPUs available in the cluster without loosing efficiency.

Wide-Area Phase-Angle Measurements for Islanding Detection—An Adaptive Nonlinear Approach

The integration of an ever growing proportion of large scale distributed renewable generation has increased the probability of maloperation of the traditional RoCoF and vector shift relays. With reduced inertia due to non-synchronous penetration in a power grid, system wide disturbances have forced the utility industry to design advanced protection schemes to prevent system degradation and avoid cascading outages leading to widespread blackouts. This paper explores a novel adaptive nonlinear approach applied to islanding detection, based on wide area phase angle measurements. This is challenging, since the voltage phase angles from different locations exhibit not only strong nonlinear but also time-varying characteristics. The adaptive nonlinear technique, called moving window kernel principal component analysis is proposed to model the time-varying and nonlinear trends in the voltage phase angle data. The effectiveness of the technique is exemplified using both DigSilent simulated cases and real test cases recorded from the Great Britain and Ireland power systems by the OpenPMU project.

Modelling the fresh properties of self-compacting concrete using support vector machine approach

The main objective of the study presented in this paper was to investigate the feasibility using support vector machines (SVM) for the prediction of the fresh properties of self-compacting concrete. The radial basis function (RBF) and polynomial kernels were used to predict these properties as a function of the content of mix components. The fresh properties were assessed with the slump flow, T50, T60, V-funnel time, Orimet time, and blocking ratio (L-box). The retention of these tests was also measured at 30 and 60 min after adding the first water. The water dosage varied from 188 to 208 L/m3, the dosage of superplasticiser (SP) from 3.8 to 5.8 kg/m3, and the volume of coarse aggregates from 220 to 360 L/m3. In total, twenty mixes were used to measure the fresh state properties with different mixture compositions. RBF kernel was more accurate compared to polynomial kernel based support vector machines with a root mean square error (RMSE) of 26.9 (correlation coefficient of R2 = 0.974) for slump flow prediction, a RMSE of 0.55 (R2 = 0.910) for T50 (s) prediction, a RMSE of 1.71 (R2 = 0.812) for T60 (s) prediction, a RMSE of 0.1517 (R2 = 0.990) for V-funnel time prediction, a RMSE of 3.99 (R2 = 0.976) for Orimet time prediction, and a RMSE of 0.042 (R2 = 0.988) for L-box ratio prediction, respectively. A sensitivity analysis was performed to evaluate the effects of the dosage of cement and limestone powder, the water content, the volumes of coarse aggregate and sand, the dosage of SP and the testing time on the predicted test responses. The analysis indicates that the proposed SVM RBF model can gain a high precision, which provides an alternative method for predicting the fresh properties of SCC.

On the Privacy of Encrypted Skype Communications

The privacy of voice over IP (VoIP) systems is achieved by compressing and encrypting the sampled data. This paper investigates in detail the leakage of information from Skype, a widely used VoIP application. In this research, it has been demonstrated by using the dynamic time warping (DTW) algorithm, that sentences can be identified with an accuracy of 60%. The results can be further improved by choosing specific training data. An approach involving the Kalman filter is proposed to extract the kernel of all training signals.

Hα Line Profile Asymmetries and the Chromospheric Flare Velocity Field

The asymmetries observed in the line profiles of solar flares can provide important diagnostics of the properties and dynamics of the flaring atmosphere. In this paper the evolution of the Hα and Ca ii λ8542 lines are studied using high spatial, temporal, and spectral resolution ground-based observations of an M1.1 flare obtained with the Swedish 1 m Solar Telescope. The temporal evolution of the Hα line profiles from the flare kernel shows excess emission in the red wing (red asymmetry) before flare maximum and excess in the blue wing (blue asymmetry) after maximum. However, the Ca ii λ8542 line does not follow the same pattern, showing only a weak red asymmetry during the flare. RADYN simulations are used to synthesize spectral line profiles for the flaring atmosphere, and good agreement is found with the observations. We show that the red asymmetry observed in Hα is not necessarily associated with plasma downflows, and the blue asymmetry may not be related to plasma upflows. Indeed, we conclude that the steep velocity gradients in the flaring chromosphere modify the wavelength of the central reversal in the Hα line profile. The shift in the wavelength of maximum opacity to shorter and longer wavelengths generates the red and blue asymmetries, respectively.

Face Recognition in the Scrambled Domain via Salience-Aware Ensembles of Many Kernels

With the rapid development of internet-of-things (IoT), face scrambling has been proposed for privacy protection during IoT-targeted image/video distribution. Consequently in these IoT applications, biometric verification needs to be carried out in the scrambled domain, presenting significant challenges in face recognition. Since face models become chaotic signals after scrambling/encryption, a typical solution is to utilize traditional data-driven face recognition algorithms. While chaotic pattern recognition is still a challenging task, in this paper we propose a new ensemble approach – Many-Kernel Random Discriminant Analysis (MK-RDA) to discover discriminative patterns from chaotic signals. We also incorporate a salience-aware strategy into the proposed ensemble method to handle chaotic facial patterns in the scrambled domain, where random selections of features are made on semantic components via salience modelling. In our experiments, the proposed MK-RDA was tested rigorously on three human face datasets: the ORL face dataset, the PIE face dataset and the PUBFIG wild face dataset. The experimental results successfully demonstrate that the proposed scheme can effectively handle chaotic signals and significantly improve the recognition accuracy, making our method a promising candidate for secure biometric verification in emerging IoT applications.