The present aim was to compare the accuracy of several algorithms in classifying data collected from food scent samples. Measurements using an electronic nose (eNose) can be used for classification of different scents. An eNose was used to measure scent samples from seven food scent sources, both from an open plate and a sealed jar. The k-Nearest Neighbour (k-NN) classifier provides reasonable accuracy under certain conditions and uses traditionally the Euclidean distance for measuring the similarity of samples. Therefore, it was used as a baseline distance metric for the k-NN in this paper. Its classification accuracy was compared with the accuracies of the k-NN with 66 alternative distance metrics. In addition, 18 other classifiers were tested with raw eNose data. For each classifier various parameter settings were tried and compared. Overall, 304 different classifier variations were tested, which differed from each other in at least one parameter value. The results showed that Quadratic Discriminant Analysis, MLPClassifier, C-Support Vector Classification (SVC), and several different single hidden layer Neural Networks yielded lower misclassification rates applied to the raw data than k-NN with Euclidean distance. Both MLP Classifiers and SVC yielded misclassification rates of less than 3% when applied to raw data. Furthermore, when applied both to the raw data and the data preprocessed by principal component analysis that explained at least 95% or 99% of the total variance in the raw data, Quadratic Discriminant Analysis outperformed the other classifiers. The findings of this study can be used for further algorithm development. They can also be used, for example, to improve the estimation of storage times of fruit.

Varieties of impediments expose the Mining industry to unexpected equipment failures, high commodity price damages, and various environmental difficulties and safety challenges. Therefore, we are looking for an appropriate approach to mitigate the cost of the sudden fault of Mining operations by increasing the automation level of monitoring and operating systems. To attain this goal, we utilize Machine Learning models and data fusion techniques on measured data values. As a result, we are capable of classifying the five diverse mining operations. The outcome of the method presented in this paper assists in the early fault detection of mining operations. This approach can support fault detection and prevention procedures. Besides, it leads to acceleration and enhancement of the maintenance and repair operations by simplifying the required query time.

In mining and construction operations, the protection, safety and machinery's lifetime hold a crucial concern that can impose unwelcoming costs on the projects. The motivation behind this work is to deliver a model capable of addressing these apprehensions besides managing the potential risks and costs of the types of machinery. The presented model in this article aims to increase the quality and reliability of the products and their operations by utilizing sensor information for real-time prediction and categorization of drilling operations. This model works based on the time analyses on the sensory fused data. We applied the model on the three-axis acceleration and angular velocity signals (generated from a simulated system) to extract features and categorize three different rock drilling operations. For each operation, we measured the Median Absolute Deviation (MAD) and dynamic range parameters of the acceleration signals. In addition, we succeeded to calculate the Root Mean Square (RMS) parameter as a feature from angular velocity signals. The obtained results in this study approve the real-time prediction and categorization potential of the introduced approach for the different rock drilling operations. However, the limitation of this work can be the source of the data which is originating from the simulated normal operations. As an extending future work in future publications, we will include the faulty operation data, the real data from measurements and present data analysis of abnormal operations.

A major challenge in modelling and simulation is the need to combine expertise in both software technologies and a given scientific domain. When High-Performance Computing (HPC) is required to solve a scientific problem, software development becomes a problematic issue. Considering the complexity of the software for HPC, it is useful to identify programming languages that can be used to alleviate this issue. Because the existing literature on the topic of HPC is very dispersed, we performed a Systematic Mapping Study (SMS) in the context of the European COST Action cHiPSet. This literature study maps characteristics of various programming languages for data-intensive HPC applications, including category, typical user profiles, effectiveness, and type of articles. We organised the SMS in two phases. In the first phase, relevant articles are identified employing an automated keyword-based search in eight digital libraries. This lead to an initial sample of 420 papers, which was then narrowed down in a second phase by human inspection of article abstracts, titles and keywords to 152 relevant articles published in the period 2006–2018. The analysis of these articles enabled us to identify 26 programming languages referred to in 33 of relevant articles. We compared the outcome of the mapping study with results of our questionnaire-based survey that involved 57 HPC experts. The mapping study and the survey revealed that the desired features of programming languages for data-intensive HPC applications are portability, performance and usability. Furthermore, we observed that the majority of the programming languages used in the context of data-intensive HPC applications are text-based general-purpose programming languages. Typically these have a steep learning curve, which makes them difficult to adopt. We believe that the outcome of this study will inspire future research and development in programming languages for data-intensive HPC applications.

While the data mining intermediaries play a critical role in the rock drilling industry, they also tend to provide an optimized real-time model for the drilling systems. In addition, proper online tool condition monitoring (OTOM) methods can improve the drilling performance by accessing real-time data. Hence, OTOM methods assist depreciating error and detect unspecified faults at early stages. In this study, we proposed appropriate OTOM algorithms to develop and enhance the quality of real-time systems and provide a solution to detect and categorize various stages of drilling operation with the aid of vibration signals (especially in terms of acceleration or velocity). In particular, the proposed methods in this article perform based on statistical approaches. Therefore, in order to recognize the drilling stages, we measured the Root Mean Square (RMS) values corresponding to the acceleration signals. In the meantime, we also succeeded to distinguish the drilling stages by employing estimated power spectral density (PSD) in the frequency domain. The acquired results in this publication confirm the real-time prediction and classification potential of the proposed methods for the different drilling stages and especially for the rock drilling engineering.

Fifth generation mobile networks (5G) will rather supplement than replace current 4G networks by dramatically improving their bandwidth, capacity and reliability. This way, much more demanding use cases that simply are not achievable with today’s networks will become reality - from home entertainment, to product manufacturing and healthcare. However, many of them rely on Internet of Things (IoT) devices equipped with low-cost transmitters and sensors that generate enormous amount of data about their environment. Therefore, due to large scale of 5G systems, combined with their inherent complexity and heterogeneity, Big Data and analysis techniques are considered as one of the main enablers of future mobile networks. In this work, we recognize 5G use cases from various application domains and list the basic requirements for their development and realization.

Ground Penetrating Radar (GPR) is generally used as a non-destructive method of inspection for structures and for finding defects in concrete slabs. In this paper, GPR is used in the detection of water inside the cavities of concrete hollow core slabs. We propose an algorithm that determines the water level inside the concrete slab by analyzing the time delays of the reflections originating from inside the cavity. The algorithm is based on utilizing prior knowledge about the geometry of the hollow core slab. The presence of water was successfully detected and an estimate for the height of the water surface was obtained with a GPR system operating with a central frequency of 2.7 GHz. Based on the experiments, the proposed method holds promise in providing a robust and accurate method for the detection of water inside the concrete slabs. Results, possible future research and analysis of the feasibility of GPR systems in water detection are presented and discussed.

Genetic circuits change the status quo of cellular processes when their protein numbers cross thresholds. We investigate the regulation of RNA and protein threshold crossing propensities in Escherichia coli. From in vivo single RNA time-lapse microscopy data from multiple promoters, mutants, induction schemes and media, we study the asymmetry and tailedness (quantified by the skewness and kurtosis, respectively) of the distributions of time intervals between transcription events. We find that higher thresholds can be reached by increasing the skewness and kurtosis, which is shown to be achievable without affecting mean and coefficient of variation, by regulating the rate-limiting steps in transcription initiation. Also, they propagate to the skewness and kurtosis of the distributions of protein expression levels in cell populations. The results suggest that the asymmetry and tailedness of RNA and protein numbers in cell populations, by controlling the propensity for threshold crossing, and due to being sequence dependent and subject to regulation, may be key regulatory variables of decision-making processes in E. coli.

Measurements from an ion mobility spectrometry electronic nose (eNose) can be used for distinguishing different rooms in indoor localisation. An earlier study showed that the Nearest Neighbour classifier with Euclidean distance for features provides reasonable accuracy under certain conditions. In this paper 66 alternative distance measures are compared to the Euclidean distance and principal component analysis (PCA) is applied to the data. PCA shows that the measurements on the various channels of the eNose are correlated and that using principal components 1, 2 and 4 increases the accuracy considerably. Furthermore, the experiments revealed three Pareto optimal distance measures that reduce the misclassification rate between 9-10% while using only 82-88% of the search time compared with Euclidean distance.

We consider the data-driven motion compensation problem in inverse synthetic aperture radar (ISAR) imaging. We present optimization-based ISAR techniques and propose improvements to the range alignment, time-window selection, autofocus, time-frequency-based image reconstruction, and cross-range scaling procedures. In experiments, the improvements reduced the computational burden and increased the image contrast by 50% at best and 28% on average in several test cases, including changing translational and rotational motion.

A gravity-referenced joint angle estimation, which acts in joint space, is proposed for multiple-degree-of-freedom hydraulic manipulators. As a novelty, the estimation pairs up inertial units across a section of an open kinematic chain. In this way, the three-axis linear accelerometers and the three-axis rate gyros provide a drift-free solution for observing the motion state of a rotary joint connecting two links without relying on the full forward kinematics. For a low-noise, low-delay estimate, the linear accelerations acting on the paired inertial units are modeled and robustly combined with the principles of complementary and Kalman filtering. In pick-and-place experiments with a serial-link manipulator on a multiton, off-road forestry vehicle, joint angle sensing error of less than ±1° was achieved in spite of the dynamic interaction between the vehicle base and the terrain. Furthermore, the kinematic modeling's ability to compensate for the nonplanar, coupled 3-D linkage motion is studied for one- and two-axis rate measurements along with Cartesian path tracking. This gives new insights from the typical planar kinematic models and heavy-duty control viewpoints.

Ranked Based Fingerprinting uses only ordering indices instead of actual Wi-Fi RSS values in order to make the algorithm insensitive to devices. A key component of the RBF algorithm is a similarity measure which is used to compare and find the closest ranked fingerprints. Previous papers study a few similarity measures; here we study 49 similarity measures in a test with a benchmark with publicly available indoor positioning database. For different similarity measures the positioning accuracy varies from 15.80 m to 55.22 m. The top 3 similarity measures are Lorentzian, Hamming and Jaccard. Hamming and Jaccard similarity measures have been studied in other papers while Lorenzian had not been studied with that kind of problems.

Multichannel synthetic aperture radar (MSAR) systems are essential for applications such as ground moving target indication (GMTI), interferometric SAR (InSAR), and high-resolution wide-swath (HRWS) imaging. In this paper, we analyze and compare MSAR image reconstruction algorithms. Previously, image reconstruction for MSAR has relied heavily on frequency domain matched filtering. Time domain image reconstruction algorithms have several attractive qualities, but their use has been limited due to a high computational burden. In this paper, we utilize digital beamforming and the phase center approximation to develop a fast time domain (fast factorized back-projection, FFBP) algorithm for MSAR. We present two FFBP implementations for MSAR and perform a comparative study between MSAR imaging algorithms. The numerical results confirm the feasibility of the proposed FFBP algorithms for MSAR.

Asymmetries in the distribution of time intervals between consecutive RNA productions from a gene can play a critical role in, e.g., allowing/preventing the RNA and, thus, protein numbers to cross thresholds involved in gene network decision making. Here, we use a stochastic, multi-step model of transcription initiation, with all rate constants empirically validated, and explore how the kinetics of its steps affect the temporal asymmetries in RNA production, as measured by the skewness of the distribution of intervals between consecutive RNA productions in individual cells. From the model, first, we show that this skewness differs widely with the mean fraction of time that the RNA polymerase spends in the steps preceding open complex formation, while being independent of the mean transcription rate. Next, we provide empirical validation of these results, using qPCR and live, time-lapse, single-molecule RNA microscopy measurements of the transcription kinetics of multiple promoters. We conclude that the skewness in RNA production kinetics is subject to regulation by the kinetics of the steps in transcription initiation and, thus, evolvable.

To address the tip positioning problem among flexible link robotic applications, we propose a new circular beam deformation model that is ideally suited for networked linear accelerometers and rate gyros. Importantly, the tip positioning problem of multiple-flexible-link system is reduced to the tip positioning problem of an individual flexible link subject to the circularly characterized deformation. The validity of the theoretical results is verified by experiments with a highly flexible single-link hydraulic manipulator.

Student commitment is a major concern for universities around the world. Research has indicated that students' psychological attachment to their university, in other words commitment, can be a major predictor of student retention, as well as affecting many other attitudes and types of behavior. This makes university commitment critical to university success. The present study seeks to find out whether there is a relation between students' grades and the degree of their engagement and commitment, using the Evolute approach. The research group in this study consisted of master's degree students studying in faculties of Business and Built Environment and Computing and Electrical Engineering. In the newly developed instrument, respondents assessed 124 unique statements regarding their current situation and their vision for the future. This score and its sub-scores are compared in relation to individual course grades.

The objective of the radar performance prediction is to compute quantities of interest concerning the ability of the radar to observe its surroundings. Besides the properties of the radar system, the performance is affected by the target, whose radar cross section (RCS) is one of the predominant Factors. The performance prediction is usually performed in relation to the target RCS characterized by a constant value or a particular statistical distribution. Such representations generalize real-life complex targets rendering them unsuitable for some objectives since the RCS is significantly influenced by the target aspect angle and is inherently stochastic by nature. Thus, a more dynamic description may be valuable e.g. for analyzing the radar performance on a flight path of interest. We propose representing the RCS with a histogram that includes such dynamic properties and is suitable for considering the target in different ways for performance prediction: in a more general manner or dependent on its aspect angle. We consider the case of traditional RCS with low spatial resolution and demonstrate the proposed approach through the probability of detection computed for a generic surveillance radar.

We present a new quantitative method of estimation of fluorescent molecule numbers from time-lapse, single-cell, fluorescence microscopy data. Its main aim is to eradicate backward propagation of noise, which is present in previous methods. The method is first validated using Monte Carlo simulations. These tests show that when the time-lapse data are corrupted with negative noise, the method obtains significantly more precise results than current techniques. The applicability of the method is demonstrated on novel time-lapse, single-cell measurements of fluorescently tagged ribonucleic acid (RNA) molecules. Interestingly, we find that the intervals inferred by the new method have the same mean but reduced variability when compared to the previously existing method, which, in accordance to human observers, is a more accurate estimation.