Search

This paper deals with the Adaptive neuro-fuzzy inference system (ANFIS)–based load frequency controller (LFC). These controllers are projected for load frequency control of thermal-Photovoltaic (PV) power generation entity as a hybrid power system. In this study, random solar isolation is applied to the proposed hybrid power system. The proposed hybrid power system consists of a PV power unit with a maximum power point tracking control, a PV inverter, and an AC load. Simulations are performed with structural change in the load setting. The solar isolation results are compared with conventional proportional-integral-derivative (PID) and fuzzy logic controller (FLC). The results are then projected with an ANFIS based LFC. The simulation results observed that ANFIS attains a relatively better response for the frequency deviation profile. It typically controls the frequency deviation of a given hybrid power system and thereby advances the dynamic performances. The results also show that the performance of the hybrid power system with the use of ANFIS based neuro-fuzzy controllers attains relatively better than those which attains by the PID and FLC.

Since the beginning of the coronavirus disease (COVID-19) pandemic several machine learning and deep learning methods had been introduced to detect the infected patients using the X-Ray or CT scan images. Numerous sophisticated data-driven methods had been introduced to improve the performance and the accuracy of the diagnosis models. This paper proposes an improved densely connected convolutional networks (DenseNet) method based on transfer learning (TL) to enhance the model performance. The results show promising model accuracy.

Advances in artificial intelligence, smart sensors, data mining, and other fields of ICT have resulted in a plethora of research projects aimed at harnessing these technologies, for example to generate new knowledge about diseases, to develop systems for better management of chronic diseases, and to assist the elderly with independent living. While the algorithms themselves can be developed using anonymized or synthetic data, conducting a pilot study is often one of the key components of a research project, and such studies unavoidably involve actual users with their personal data. Although one of the derogations stipulated in Article 89 of the GDPR is related to the data processed for scientific purposes, the GDPR still is applicable to that processing in a broader interpretation. The computer scientists and engineers working in research projects may not always be fully familiar with all the details of the GDPR, a close collaboration with a lawyer specialized in the European data protection legislation is highly beneficial for the success of a project. In this paper, we consider a hypothetical research project developed by an engineer dealing with sensitive personal data and a lawyer conducting Data Protection Impact Assessment to ensure legality and quality of the research project. Povzetek: Prispevek obravnava oceno ucinka v zvezi z varstvom podatkov pri hipoteticnem raziskovalnem projektu, pri katerem se z metodami umetne inteligence analizira medicinske podatke uporabnikov.

An accurate outbreak prediction of COVID-19 can successfully help to get insight into the spread and consequences of infectious diseases. Recently, machine learning (ML) based prediction models have been successfully employed for the prediction of the disease outbreak. The present study aimed to engage an artificial neural network-integrated by grey wolf optimizer for COVID-19 outbreak predictions by employing the Global dataset. Training and testing processes have been performed by time-series data related to January 22 to September 15, 2020 and validation has been performed by time-series data related to September 16 to October 15, 2020. Results have been evaluated by employing mean absolute percentage error (MAPE) and correlation coefficient (r) values. ANN-GWO provided a MAPE of 6.23, 13.15 and 11.4% for training, testing and validating phases, respectively. According to the results, the developed model could successfully cope with the prediction task.

An accurate outbreak prediction of COVID-19 can successfully help to get insight into the spread and consequences of infectious diseases. Recently, machine learning (ML) based prediction models have been successfully employed for the prediction of the disease outbreak. The present study aimed to engage an artificial neural network-integrated by grey wolf optimizer for COVID-19 outbreak predictions by employing the Global dataset. Training and testing processes have been performed by time-series data related to January 22 to September 15, 2020 and validation has been performed by time-series data related to September 16 to October 15, 2020. Results have been evaluated by employing mean absolute percentage error (MAPE) and correlation coefficient (r) values. ANN-GWO provided a MAPE of 6.23, 13.15 and 11.4% for training, testing and validating phases, respectively. According to the results, the developed model could successfully cope with the prediction task.

Nowadays, Machine Learning applications are spreading widely in different science and research fields which gave, in fact, the possibility to enhance the results of all kind of both, automated tasks and further possible application areas. Autonomous smart driving cars presents one of the major fields that uses machine learning technics to push further the automated tasks inside the car systems. Many types of research related to this topic enabled real application to fully automate some parts of the car driving process. Road lane detection, pedestrian and car approximation detection, and fastest road finding using real-time traffic statistics present some of the possible application areas that could use the Machine learning technics to improve autonomous driving cars systems. Traffic signs present an important part of the daily driving routine, therefore, traffic signs recognition for mobile-based application is a great solution that provides a new layer for autonomous car driving systems. In this paper, we propose a powerful tool for traffic signs recognition in a mobile-based application. This tool uses TensorFlow together with transfer learning technic that makes it easier to train our dataset on a pre-trained Model using the convolutional network (ConvNet). The used model is a Single Shot MultiBox Detector (SSD) MobileNet V2 based model which uses one single deep network to train the model on multiple objects per image. This network uses 300x300 annotated input images with multiple objects to provide faster training time and faster detection results compared to other types of neural networks. The annotation is made by providing the coordinates of the rectangle that surrounds the given object together with its label which defines the name of the object. The coordinates are usually given by providing the (x,y) coordinates of the top-left and bottom-right points of the surrounding rectangle. This presents a powerful technic for real-time detection on mobile devices with low computational capabilities. The resulting model of the training is then converted to a TensorFlow Lite quantized model using TensorFlow Lite converter which provides compatibility with mobile devices with low computational capacity. The quantized model showed 4 times faster detection compared to the float model on the mobile device.

In this paper an accessible indoor navigation data model suitable for multi-model NoSQL databases is proposed, based on iLOC and iACC ontologies. We consider our previous observations about practicality and performance of labeled property graph data models used for accessible indoor wayfinding and suggest multiple ways of improvement by using a multi-model approach. A very important aim of the new data model is to provide a structure to be able to run fully customizable wayfinding queries with respect to accessibility needs of disabled users.

Several outbreak prediction models for COVID-19 are being used by officials around the world to make informed decisions and enforce relevant control measures. Among the standard models for COVID-19 global pandemic prediction, simple epidemiological and statistical models have received more attention by authorities, and these models are popular in the media. Due to a high level of uncertainty and lack of essential data, standard models have shown low accuracy for long-term prediction. Although the literature includes several attempts to address this issue, the essential generalization and robustness abilities of existing models need to be improved. This paper presents a comparative analysis of machine learning and soft computing models to predict the COVID-19 outbreak as an alternative to susceptible&ndash, infected&ndash, recovered (SIR) and susceptible-exposed-infectious-removed (SEIR) models. Among a wide range of machine learning models investigated, two models showed promising results (i.e., multi-layered perceptron, MLP, and adaptive network-based fuzzy inference system, ANFIS). Based on the results reported here, and due to the highly complex nature of the COVID-19 outbreak and variation in its behavior across nations, this study suggests machine learning as an effective tool to model the outbreak. This paper provides an initial benchmarking to demonstrate the potential of machine learning for future research. This paper further suggests that a genuine novelty in outbreak prediction can be realized by integrating machine learning and SEIR models.

Prediction models in mobility and transportation maintenance systems have been dramatically improved through using machine learning methods. This paper proposes novel machine learning models for intelligent road inspection. The traditional road inspection systems based on the pavement condition index (PCI) are often associated with the critical safety, energy and cost issues. Alternatively, the proposed models utilize surface deflection data from falling weight deflectometer (FWD) tests to predict the PCI. Machine learning methods are the single multi-layer perceptron (MLP) and radial basis function (RBF) neural networks as well as their hybrids, i.e., Levenberg-Marquardt (MLP-LM), scaled conjugate gradient (MLP-SCG), imperialist competitive (RBF-ICA), and genetic algorithms (RBF-GA). Furthermore, the committee machine intelligent systems (CMIS) method was adopted to combine the results and improve the accuracy of the modeling. The results of the analysis have been verified through using four criteria of average percent relative error (APRE), average absolute percent relative error (AAPRE), root mean square error (RMSE), and standard error (SD). The CMIS model outperforms other models with the promising results of APRE=2.3303, AAPRE=11.6768, RMSE=12.0056, and SD=0.0210., 23 pages, 10 figures

The importance of energy systems and their role in economics and politics is not hidden for anyone. This issue is not only important for the advanced industrialized countries, which are major energy consumers but is also essential for oil-rich countries. In addition to the nature of these fuels, which contains polluting substances, the issue of their ending up has aggravated the growing concern. Biofuels can be used in different fields for energy production like electricity production, power production, or for transportation. Various scenarios have been written about the estimated biofuels from different sources in the future energy system. The availability of biofuels for the electricity market, heating, and liquid fuels is critical. Accordingly, the need for handling, modeling, decision making, and forecasting for biofuels can be of utmost importance. Recently, machine learning (ML) and deep learning (DL) techniques have been accessible in modeling, optimizing, and handling biodiesel production, consumption, and environmental impacts. The main aim of this study is to review and evaluate ML and DL techniques and their applications in handling biofuels production, consumption, and environmental impacts, both for modeling and optimization purposes. Hybrid and ensemble ML methods, as well as DL methods, have found to provide higher performance and accuracy.

Automated controlling the harvesting systems can significantly increase the efficiency of the agricultural practices and prevent food wastes. Modeling and improvement of the combine harvester can increase the overall performance. Machine learning methods provide the opportunity of advanced modeling for accurate prediction of the highest performance of the machine. In this study, the modeling of combine harvesting id performed using radial basis function (RBF) and the hybrid machine learning method of adaptive neuro-fuzzy inference system (ANFIS) to predict various variables of the combine harvester for the optimal performance. Response surface methodology (RSM) is also used to optimize the models. The comparative study shows that the ANFIS method outperforms the RBF method.

Artificial intelligence methods and application have recently shown great contribution in modeling and prediction of the hydrological processes, climate change, and earth systems. Among them, deep learning and machine learning methods mainly have reported being essential for achieving higher accuracy, robustness, efficiency, computation cost, and overall model performance. This paper presents the state of the art of machine learning and deep learning methods and applications in this realm and the current state, and future trends are discussed. The survey of the advances in machine learning and deep learning are presented through a novel classification of methods. The paper concludes that deep learning is still in the first stages of development, and the research is still progressing. On the other hand, machine learning methods are already established in the fields, and novel methods with higher performance are emerging through ensemble techniques and hybridization.

The conventional machine learning (ML) algorithms are continuously advancing and evolving at a fast-paced by introducing the novel learning algorithms. ML models are continually improving using hybridization and ensemble techniques to empower computation, functionality, robustness, and accuracy aspects of modeling. Currently, numerous hybrid and ensemble ML models have been introduced. However, they have not been surveyed in a comprehensive manner. This paper presents the state of the art of novel ML models and their performance and application domains through a novel taxonomy.

Deep learning (DL) algorithms have recently emerged from machine learning and soft computing techniques. Since then, several deep learning (DL) algorithms have been recently introduced to scientific communities and are applied in various application domains. Today the usage of DL has become essential due to their intelligence, efficient learning, accuracy and robustness in model building. However, in the scientific literature, a comprehensive list of DL algorithms has not been introduced yet. This paper provides a list of the most popular DL algorithms, along with their applications domains.

Building energy consumption plays an essential role in urban sustainability. The prediction of the energy demand is also of particular importance for developing smart cities and urban planning. Machine learning has recently contributed to the advancement of methods and technologies to predict demand and consumption for building energy systems. This paper presents a state of the art of machine learning models and evaluates the performance of these models. Through a systematic review and a comprehensive taxonomy, the advances of machine learning are carefully investigated and promising models are introduced.

Automated controlling the harvesting systems can significantly increase the efficiency of the agricultural practices and prevent food wastes. Modeling and improvement of the combine harvester can increase the overall performance. Machine learning methods provide the opportunity of advanced modeling for accurate prediction of the highest performance of the machine. In this study, the modeling of combine harvesting id performed using radial basis function (RBF) and the hybrid machine learning method of adaptive neuro-fuzzy inference system (ANFIS) to predict various variables of the combine harvester for the optimal performance. Response surface methodology (RSM) is also used to optimize the models. The comparative study shows that the ANFIS method outperforms the RBF method.

The conventional machine learning (ML) algorithms are continuously advancing and evolving at a fast-paced by introducing the novel learning algorithms. ML models are continually improving using hybridization and ensemble techniques to empower computation, functionality, robustness, and accuracy aspects of modeling. Currently, numerous hybrid and ensemble ML models have been introduced. However, they have not been surveyed in a comprehensive manner. This paper presents the state of the art of novel ML models and their performance and application domains through a novel taxonomy.

Building energy consumption plays an essential role in urban sustainability. The prediction of the energy demand is also of particular importance for developing smart cities and urban planning. Machine learning has recently contributed to the advancement of methods and technologies to predict demand and consumption for building energy systems. This paper presents a state of the art of machine learning models and evaluates the performance of these models. Through a systematic review and a comprehensive taxonomy, the advances of machine learning are carefully investigated and promising models are introduced.

The importance of energy systems and their role in economics and politics is not hidden for anyone. This issue is not only important for the advanced industrialized countries, which are major energy consumers but is also essential for oil-rich countries. In addition to the nature of these fuels, which contains polluting substances, the issue of their ending up has aggravated the growing concern. Biofuels can be used in different fields for energy production like electricity production, power production, or for transportation. Various scenarios have been written about the estimated biofuels from different sources in the future energy system. The availability of biofuels for the electricity market, heating, and liquid fuels is critical. Accordingly, the need for handling, modeling, decision making, and forecasting for biofuels can be of utmost importance. Recently, machine learning (ML) and deep learning (DL) techniques have been accessible in modeling, optimizing, and handling biodiesel production, consumption, and environmental impacts. The main aim of this study is to review and evaluate ML and DL techniques and their applications in handling biofuels production, consumption, and environmental impacts, both for modeling and optimization purposes. Hybrid and ensemble ML methods, as well as DL methods, have found to provide higher performance and accuracy.

Artificial intelligence methods and application have recently shown great contribution in modeling and prediction of the hydrological processes, climate change, and earth systems. Among them, deep learning and machine learning methods mainly have reported being essential for achieving higher accuracy, robustness, efficiency, computation cost, and overall model performance. This paper presents the state of the art of machine learning and deep learning methods and applications in this realm and the current state, and future trends are discussed. The survey of the advances in machine learning and deep learning are presented through a novel classification of methods. The paper concludes that deep learning is still in the first stages of development, and the research is still progressing. On the other hand, machine learning methods are already established in the fields, and novel methods with higher performance are emerging through ensemble techniques and hybridization.

Deep learning (DL) algorithms have recently emerged from machine learning and soft computing techniques. Since then, several deep learning (DL) algorithms have been recently introduced to scientific communities and are applied in various application domains. Today the usage of DL has become essential due to their intelligence, efficient learning, accuracy and robustness in model building. However, in the scientific literature, a comprehensive list of DL algorithms has not been introduced yet. This paper provides a list of the most popular DL algorithms, along with their applications domains.

Emotion detection from the text is an important and challenging problem in text analytics. The opinion-mining experts are focusing on the development of emotion detection applications as they have received considerable attention of online community including users and business organization for collecting and interpreting public emotions. However, most of the existing works on emotion detection used less efficient machine learning classifiers with limited datasets, resulting in performance degradation. To overcome this issue, this work aims at the evaluation of the performance of different machine learning classifiers on a benchmark emotion dataset. The experimental results show the performance of different machine learning classifiers in terms of different evaluation metrics like precision, recall ad f-measure. Finally, a classifier with the best performance is recommended for the emotion classification., 30 pages, 13 tables, 1 figure

Functional dependency detection is a very useful tool for many fields of computer science, e.g. machine learning. The presence of functional dependency in a dataset means that there is redundancy on the attribute level, so the dataset can be cropped without any loss of classification efficiency, thus making the training process faster. Many algorithms have been proposed for the detection and extraction of functional dependencies, however, they generally do not scale well with the amount of data. In this paper, a functional dependency detection method is proposed that is linear in the number of data tuples. It is based on Sequential Fuzzy Indexing Tables, a classifier structure that uses a significant amount of memory in return for a very fast operation.

Machine learning (ML) models have been widely used in diverse applications of energy systems such as design, modeling, complex mappings, system identification, performance prediction, and load forecasting. In particular, the last two decades has seen a dramatic increase in the development and application of various types of ML models for energy systems. This paper presents the state of the art of ML models used in energy systems along with a taxonomy of applications and methods. Through a novel search methodology, ML models are identified and further classified according to the ML modeling technique, energy type, and the application area. Furthermore, a comprehensive review of the literature represents an assessment and performance evaluation of the ML models, their applications and a discussion on the major challenges and opportunities for prospective research. This paper further concludes that there is an outstanding rise in the accuracy, robustness, precision and the generalization ability of the ML models in energy systems using the hybrid and ensemble ML algorithms. ML models are widely used in solar energy and wind energy prediction so that these sustainable energy sources will become more practical and more economic. Energy demand prediction by ML models will also improve our communities’ sustainability. &nbsp

The free-climbing robots are the alternative to wheeled vehicles, because the legged locomotion allows mobility on the steep gradients under the harsh environmental conditions. The steady progression requires coordination of leg movements for which the force sensors implemented in the joints of the robot legs are not always sufficient. Considering the previous developments in the field of robotics, the combination of the impedance and vision sensors comprises an obvious solution for providing stable and accurate contact between the individual legs and the ground. The implementation of additional, non-invasive sensor/s on all the legs of the hexapod robotic platform, for the indirect measuring and analysis of the terrain surface, is investigated. Furthermore, a method for the parametric 3D modeling of simulation assemblies is described in a mixed top-down, bottom-up and middle-out design approach. Mechanisms were created using the common origin skeletal modeling (top-down approach), allowing the context analysis and the definition of the parameters for the further numerical optimization in an early phase of the development. Finally, the applicability of the anytime approach during the optimization of the mechanical parts of the mechanism in the design process was tested.

In this paper, we present a Cluster-Based Approach (CBA) that utilizes the support vector machine (SVM) and an artificial neural network (ANN) to estimate and predict the daily horizontal global solar radiation. In the proposed CBA-ANN-SVM approach, we first conduct clustering analysis and divided the global solar radiation data into clusters, according to the calendar months. Our approach aims at maximizing the homogeneity of data within the clusters, and the heterogeneity between the clusters. The proposed CBA-ANN-SVM approach is validated and the precision is compared with ANN and SVM techniques. The mean absolute percentage error (MAPE) for the proposed approach was reported lower than those of ANN and SVM.

Nowadays, parallelization is an increasingly popular tool to speed up algorithms. Data classification is one of the many fields of computer science that can take significant advantage of that. In this paper, a parallel implementation of Fuzzy RBF based filters are proposed for pattern recognition problems. It realizes a simple pattern matching by using the radial basis functions for proximity detection, then simply choosing the class or label associated to the pattern as output. The classifier has the advantage of being very simple to implement, to train and to modify the obtained knowledge. With the parallel computing improvement, the speed of both the training and evaluation phase are significantly increased compared to the sequential implementation.

With the prevalence of smart homes, systems that realize the Ambient Assisted Living paradigm are getting increasingly common. This paper presents an AAL application that aims to assist its users with a sedentary Iifestyle improving their health by suggesting gradual changes in their diet that align with their personal tastes, and by warning them if they had not taken a break from sitting by their computer or workstation for a given period of time. The former is a high complexity optimization problem, as the task is to achieve a diet in the long term that balances the daily intake of dozens of important nutrients, while making sure that the changes align with the personal taste and dietary restrictions (e.g., celiac disease) of the user.

Nowadays, the changes in people’s habits and the development of technology result in the wide spreading of the model-based health monitoring systems. In this paper, an improvement of the measurement evaluation method is introduced, which is used during the sport activity in real-time. The basis of the novel approach is our previously reported anytime hierarchical fuzzy risk calculation model, which is able to handle some uncertainties, imprecision, and subjectivity in the data and in the evaluation process and can cope with the dynamically changing environment, available time, and resources. In the new model the input membership functions, which are tuned according to the patient characteristics, are modified based on the data recorded during previous measurements under the same conditions. By this, the person-dependent characteristics and the unavoidable changing of the dynamic reactions of the human organism can also be considered and the risk level can be more reliably predicted.

In this paper, a new filter network is presented that is based on Radial Base Function Networks (RBFNs). The output layer of the network is modified, in order to make it more effective in certain fuzzy control systems. The training of the network is solved by a clustering step, for which two different clustering methods are proposed. The suggested structure can efficiently be used for data classification.

This chapter presents an easily implementable method of fuzzy shape extraction for shape recognition. The method uses Fuzzy Hypermatrix-based classifiers in order to find the potential location of the target objects based on their colors, then determines the areas where the most densely occurring positive findings in order to restrict the area of operation thus speeding the process up. In these areas the edges are detected, the edges are mapped to tree structures, which are trimmed down to simple outline sequences using heuristics from the Fuzzy Hypermatrix. Finally, fuzzy information is extracted from the outlines that can be used to classify the shape with a fuzzy inference machine.

Lyapunov’s 2nd or Direct method is recognized as being the primary tool of adaptive control of nonlinear dynamic systems. The great majority of the adaptive nonlinear control design rest on Lyapunov’s stability theorem. Recent findings have revealed that the Robust Fixed Point Transformation-based method can succesfully replace the Lyapunov technique. Later the “Sigmoid Generated Fixed Point Transformation (SGFPT)” has been introduced. This systematic method has been proposed for the generation of whole families of Fixed Point Transformations. Its extension from Single Input Single Output (SISO) to Multiple Input Multiple Output (MIMO) systems has also been given. In recent times, the great majority of model building issues are replaced by “Soft Computing” techniques. In contrast to the classical mathematical methods the intelligent methodologies are able to cope with ill-defined systems, disturbances and missing information by an efficient and robust way. Especially fuzzy logic has become to be used to model complex systems. This contribution makes an attempt to utilize the advantages of fuzzy approximation in the SGFPT control design. The theoretical investigations are validated by the adaptive control of the inverted pendulum. Comparative analysis have been carried out between the “affine” and the “soft computing-based” models. Results of numerical simulations confirm the applicability and efficiency of the proposed method.

Nowadays, there are numerous types of diets that aim to improve the quality of life, health and longevity of people. However, these diets typically involve a strictly planned regime, which can be hard to get used to or even to follow through at all, due to the sudden nature of the change. In this paper, the framework for an Intelligent Space application is proposed that helps its users to achieve a healthier diet in the long term by introducing small, gradual changes into their consumption habits. The application observes the daily nutrition intake of its users, applies data mining in order to learn their personal tastes, and educates them about the effects of their current diet on their health. Then it analyzes the knowledge base to find different food or drink items that align with the perceived preferences, while also add to the balance of the daily nutrition of the users considering their physical properties, activities, and health conditions (e.g. diabetes, celiac disease, food allergies, etc). Finally, the system uses the findings to make suggestions about adding items from the consumption list, or change one item to another.

In railway system dynamics the dynamic stability problem has significant role particularly when it comes to dealing with the motion of the vertically deformable joints on damped Winkler foundation. Timoshenko and Euler-Bernoulli beam equations are the two widely used methods for dynamics analysis of this problem. This paper describes a comparison between Euler-Bernoulli and Timoshenko beam equations to investigate the track motion dynamic stability via solving the fourth order partial differential of the both models on an Elastic Foundation. This article aims at identifying an efficient model for future investigation on the track motion dynamics stability for the advanced railway systems.

The multiobjective optimization (MOO) software package of modeFrontier has recently become popular within industries, academics and research communities. Today, universities as well as research institutes are using modeFrontier optimization toolboxes for teaching and research proposes around the world. One of the reason behind the popularity of the package, is the way it utilizes the available resources in an efficient and integrated manner and providing multidimensional post-processing tools. The user-friendly design optimization environment of modeFrontier integrates various optimization methods with the major computer aided engineering codes and commercial numerical analysis tools. Among the wide range of applications of modeFrontier, the energy sector, particularly, has been highly benefiting from the advancements in design optimization. This article presents the state of the art survey of the novel applications of modeFrontier in this realm.

Utilizing dynamic resource allocation for load balancing is considered as an important optimization process of task scheduling in cloud computing. A poor scheduling policy may overload certain virtual machines while remaining virtual machines are idle. Accordingly, this paper proposes a hybrid load balancing algorithm with combination of Teaching-Learning-Based Optimization (TLBO) and Grey Wolves Optimization algorithms (GWO), which can well contribute in maximizing the throughput using well balanced load across virtual machines and overcome the problem of trap into local optimum. The hybrid algorithm is benchmarked on eleven test functions and a comparative study is conducted to verify the results with particle swarm optimization (PSO), Biogeography-based optimization (BBO), and GWO. To evaluate the performance of the proposed algorithm for load balancing, the hybrid algorithm is simulated and the experimental results are presented.

Variable length data classification is an important field of machine learning. However, while there are plenty of classifiers in literature that can efficiently handle fixed length data, not many can also handle data with varying length samples. In this paper, a structure is proposed for quick and robust classification of such data, as well as data sets with occasionally missing values. It builds on the principle of look-up table classifiers, realizing a direct assignment between the attribute values of the given data samples and their corresponding classes. The proposed data structure solves this problem by decomposing the problem space into a sequence of integer value combinations, thus creating and maintaining a layered structure in the combined form of 1D and 2D arrays. Furthermore, a simple analysis regarding the data structure can reveal functional dependencies considering the attributes of the data set, offering an option to simplify the structure thus reduce its complexity.

Recently the intelligent space applications have become increasingly beneficial considering robot control. In this paper the visual controlling concept is presented in the iSpace framework. The positions of the end-effector of the robot manipulator are presented by the 3D spatial coordinates extracted from image pairs. The exact image Jacobian matrix of the mapping from Cartesian space to image space is given, the task space controllers can be directly extended to image-space controllers. The Jacobian matrix poses uncertainty if modeling and calibration errors are present. Despite the fact that much progress has been presented in the literature of visual servoing, there are only a few results obtained for the stability analysis in presence of the uncertain camera parameters. This research aims developing a new method for the control of the manipulator in Cartesian space, using the vision information of the environment obtained by cameras using the OptiTrack framework. The robotic manipulator is mounted on a mobile tank. The control scheme allows the end effector to transit smoothly from Cartesian-space feedback to vision-space feedback when the target is inside the vicinity of the camera. Key points on the manipulator are marked which are detected by the camera system. The framework calculates the coordinates of the markers, and thus estimate the state of each joint of the manipulator within a margin of error. In order to achieve the most precise estimation each camera image is weighted during the evaluation. The weights are determined using data set of images. After, a likelihood function is assigned for each joint that is used for defining the position and designing the motion. During the experiments the proposed control concept has proven to be reliable.

Big data analytics has become an important tool for the progress and success of a wide range of businesses and industries. Its diversity and flexibility offer a steady increasing scope for the several applications to stay competitive in the market. For that, big data approach provides several advantages such as advanced analytics, intelligent optimization, informed decision making, large-scale modeling, and accurate predictions. Due to the numerous advantages, it has been particularly possible to find more accurate and feasible solutions for the current engineering problems. Hence, the impact of big-data analytics in the engineering realm and applications is increasing more than ever. This article presents a survey to investigate how engineering community has adopted big data technologies to stay competitive. To conduct the investigation a state of the art survey of the academic literature on the big data applications to engineering is presented.

Accurate prediction models can potentially transform businesses, organizations, governments, and industries. Data-driven prediction methods and applications have recently become very popular. One of the novel method of building prediction models is to use data-driven methods and knowledge discovery on the web contents. This includes the news and media as well as social networks contents. This method uses advanced technologies of big data, machine learning, deep learning and intelligent optimization for finding patterns in big data to build prediction models. This article presents a state of the art survey on the latest technological advancements, novel methods, and applications in developing prediction models.

Recurrent Neural Networks are widely used tools for the classification of variable length data. However, their training is generally a very time-consuming task, especially for problems with high dimensions. The classification method proposed in this paper aims to provide a fast and simple alternative. Extended Sequential Fuzzy Indexing Tables are following the principle behind lookup table classifiers in that they realize an input-output association by mapping the problem space using arrays. The proposed network achieves this by breaking the multi-dimensional problem space down to a sequence of combinations, resulting in a flexible architecture that can work well with varying length data.

The goal of the research was to create online maps with dynamically changing cartographic images according to the users’ map reading skills, by only using open source tools and resources. The maps were designed for three map reader groups (beginners, intermediates and experts). We used OpenStreetMap and Shuttle Radar Topography Mission (SRTM) data to create the maps in QGIS . To display the maps online, we used QGIS Server on the server side, and OpenLayers on the client side. Based on earlier studies on map reading skills and cognitive load, the map design and generalization aimed to maximize the map reading efficiency of the target group. In the grouping process, an online test was used to measure the map reading competences of the map readers, such as the interpretation of hypsography, orientation skills, distance and travel time estimation, interpretation of map symbols, interpretation of geographic names and interpretation of topographic objects. The design and generalization of the map features were based on the map data categories (such as hypsography, coverage, roads, etc.) and they are different on the three maps.

An online map reading test was done with 859 subjects to statistically measure the efficiency of information retrieval from three different cartographic images of the same area. The differences between the maps were defined by the graphic variables of size, color, pattern, etc., for six map data categories: linear features, hydrography, land cover, elevations, point-like objects, geographic names. The subjects solved map reading tasks related to these categories. Cartographic images were designed for each of the three map reader groups: beginners, intermediates and experts. The design method and the grouping were based on the results of previous studies, and the grouping was done with a competency test prior to the map reading task. The results showed the effectiveness of information retrieval from the three different cartographic images. Conclusions about the efficiency were done concerning the age, gender and level of expertise of the subjects.

In this paper, a Sequential Fuzzy Indexing Tables classifier is proposed for problems that require fast online operation. Its base idea comes from fuzzy hypermatrices (which are specialized versions of fuzzy look-up tables) that realize nearest-neighbor classification in order to recognize patterns similar to known ones. It is done by mapping the problem space into the memory in form of multidimensional matrices, so the class of the input data can be gained instantly in the evaluation phase. The downside of the base method is that the memory requirements scale exponentially with the number of attributes and the size of the attribute domains. The proposed classifier solves this issue for problems with large, but sparse workspaces by storing only a part of the problem domain. Thus instead of using a single multidimensional matrix, the classifier consists of a layered structure, breaking the multi-dimensional problem to a sequential combination of 2D fuzzy matrices.

Chronic neuromuscular diseases often cause dysarthria (speech distortions, impaired articulation, etc.), that becomes more severe over time. This aspect of the disease represents a serious problem in voice-controlled smart home systems. Medical research suggests that some speech features are impaired considerably, while others remain relatively unharmed. Therefore, it is possible to create a distance metric based on medical data that measures difference between two speech commands in a dysarthria-specific way: the contribution of various features to the distance is based on the extent of dysarthric impairment. Specifying a minimal distance between speech commands contributes to a more effective recognition during later stages of the disease.