Search results

Toward the construction of a new paradigm in teacher education in a globalized and digitalized society where it is intended to value knowledge and teacher professional development sustained by collaboration and cooperation, training policies and models based on technology-enhanced active learning will be required. This chapter aims to analyze the dimensions that can affect these training models within a new educational paradigm, at the level of professional development and increase of technological skills, collaborative processes for the creation of communities of practice, and promotion of active learning that contribute to innovative hybrid environments and transformative learning. In the Covid-19 post-pandemic, it is crucial to study and mobilize the experiences developed in the educational field exploring how these can be harnessed to build this new educational paradigm. This work aims to contribute with a reasoned reflection and insights concerning learning models and methodologies in teacher education that contribute to transformative active learning. Focusing on the link between preservice and in-service teacher education, the interrelation among teacher education and evaluation, and the construction of innovative technology-enhanced learning environments, for instance through the active training model.

Addressing a gap in entrepreneurial training programs, the main objective of this study was to introduce a hybrid training model that provides training to entrepreneurs after they have started their operations and before they become large and/or well established. The presented model consist of a full entrepreneurship training program suited to serve entrepreneurs who have been operating for no less than 2 years, have 1–14 employees, and need basic training to further achieve their operational goals. This format allows for progressive learning while encouraging networking among participants. Using a case study, 5 years of data are presented describing this program and its value for its participants including urban entrepreneurs.

With the development of deep learning-based analytical techniques, increased research has focused on fatigue data analysis methods based on deep learning, which are gaining in popularity. However, the application of deep neural networks in the material science domain is mainly inhibited by data availability. In this paper, to overcome the difficulty of multifactor fatigue life prediction with small data sets,

A multiple neural network ensemble (MNNE) is used, and an MNNE with a general and flexible explicit function is developed to accurately quantify the complicated relationships hidden in multivariable data sets. Moreover, a variational autoencoder-based data generator is trained with small sample sets to expand the size of the training data set. A comparative study involving the proposed method and traditional models is performed. In addition, a filtering rule based on the R2 score is proposed and applied in the training process of the MNNE, and this approach has a beneficial effect on the prediction accuracy and generalization ability.

A comparative study involving the proposed method and traditional models is performed. The comparative experiment confirms that the use of hybrid data can improve the accuracy and generalization ability of the deep neural network and that the MNNE outperforms support vector machines, multilayer perceptron and deep neural network models based on the goodness of fit and robustness in the small sample case.

The experimental results imply that the proposed algorithm is a sophisticated and promising multivariate method for predicting the contact fatigue life of a coating when data availability is limited.

A data generated model based on variational autoencoder was used to make up lack of data. An MNNE method was proposed to apply in the small data case of fatigue life prediction.

The purpose of this paper is to propose a new linear-nonlinear data preprocessing-based hybrid model to achieve a more accurate result at a lower cost for wind power forecasting. For this purpose, a decomposed based series-parallel hybrid model (PKF-ARIMA-FMLP) is proposed which can model linear/nonlinear and certain/uncertain patterns in underlying data simultaneously.

To design the proposed model at first, underlying data are divided into two categories of linear and nonlinear patterns by the proposed Kalman filter (PKF) technique. Then, the linear patterns are modeled by the linear-fuzzy nonlinear series (LLFN) hybrid models to detect linearity/nonlinearity and certainty/uncertainty in underlying data simultaneously. This step is also repeated for nonlinear decomposed patterns. Therefore, the nonlinear patterns are modeled by the linear-fuzzy nonlinear series (NLFN) hybrid models. Finally, the weight of each component (e.g. KF, LLFN and NLFN) is calculated by the least square algorithm, and then the results are combined in a parallel structure. Then the linear and nonlinear patterns are modeled with the lowest cost and the highest accuracy.

The effectiveness and predictive capability of the proposed model are examined and compared with its components, based models, single models, series component combination based hybrid models, parallel component combination based hybrid models and decomposed-based single model. Numerical results show that the proposed linear-nonlinear data preprocessing-based hybrid models have been able to improve the performance of single, hybrid and single decomposed based prediction methods by approximately 66.29%, 52.10% and 38.13% for predicting wind power time series in the test data, respectively.

The combination of single linear and nonlinear models has expanded due to the theory of the existence of linear and nonlinear patterns simultaneously in real-world data. The main idea of the linear and nonlinear hybridization method is to combine the benefits of these models to identify the linear and nonlinear patterns in the data in series, parallel or series-parallel based models by reducing the limitations of the single model that leads to higher accuracy, more comprehensiveness and less risky predictions. Although the literature shows that the combination of linear and nonlinear models can improve the prediction results by detecting most of the linear and nonlinear patterns in underlying data, the investigation of linear and nonlinear patterns before entering linear and nonlinear models can improve the performance, which in no paper this separation of patterns into two classes of linear and nonlinear is considered. So by this new data preprocessing based method, the modeling error can be reduced and higher accuracy can be achieved at a lower cost.

With suitable optimization criteria, hybrid models have proven to be efficient for preparing portfolios in capital markets of developed countries. This study adapts and investigates these methods for a developing country, thus providing a novel approach to the application of banking and finance. Our specific objectives are to employ a stochastic dominance criterion to evaluate the performances of over-the-counter (OTC) companies in a developing country and to analyze them with a hybrid model involving particle swarm optimization and artificial neural networks.

In order to achieve these aims, the authors conduct a case study of OTC companies in Iran. Weekly and daily returns of 36 companies listed in this market are calculated for one year during 2014–2015. The hybrid model is particularly interesting, and the results of the study identify first-, second- and third-order stochastic dominances among these companies. The study’s chosen model uses the best performing combination of activation functions in our analysis, corresponding to TPT, where T represents hyperbolic tangent transfers and P represents linear transfers.

Our portfolios are based on the shares of companies ranked with respect to the stochastic dominance criterion. Considering the minimum and maximum numbers of shares to be 2 and 10 for each portfolio, an eight-share portfolio is determined to be optimal. Compared with the index of Iran OTC during the research period of this study, our selected portfolio achieves a significantly better performance. Moreover, the methods used in this analysis are shown to be as efficient as they were in the capital markets of developed countries.

The problem of optimizing investment portfolios has to allow for correlations among returns from the financial maintenance period under consideration if an asymmetric distribution of returns exists (Babaei et al., 2015). Therefore, it is desirable to select an appropriate criterion in order to prepare an optimal portfolio and prioritize investment options. Although a back propagation technique is very popular in artificial neural (ANN) training, it is time-consuming to train a network in this way, and other methods such as particle swarm optimization (PSO) should be considered instead. In the hybrid combination of PSO and ANN, it is not the structure of a neural network that changes. Rather, the weighting method and the training technique chosen for the network are the important aspects, and these relate to PSO, so the only role ANN plays in this process is to reduce the errors.

The hybrid model combining ANN and PSO is seen to be considerably successful for generating optimal results and appropriate activation functions. These results are consistent with the theoretical findings of Das et al. (2013) and an application of the simple PSO in a study conducted by Pederson and Chipperfield (2010). Our research results also confirm the efficiency of stochastic dominance criteria as noted in the studies conducted by Roman et al. (2013), ANN as in a study carried out by Kristijanpoller et al. (2014) and PSO as in studies conducted by Liu et al. (2015) and Deng et al. (2012). These studies were carried out in the capital markets of developed countries, whereas the authors’ analysis relates to a developing country.

The authors deduce that the tools and methods whose efficiency was proven in the capital markets of developed countries also apply to, and demonstrate efficiency in, two novel applications of portfolio optimization within developing countries. The first of these is gaining familiarity with the theory and practice of these research tools and the methods that enrich financial knowledge of investors in developing countries. The second of these is the application of tools and methods identified by investors in the capital markets of developing countries, which enables optimal allocation of financial resources and growth of the markets. The authors expect that these findings will contribute to improving the economies of developing countries and thus help with economic development and facilitation of improving trends.

Improving the accuracy and reducing computational costs of predictions, especially the prediction of time series, is one of the most critical parts of the decision-making processes and management in different areas and organizations. One of the best solutions to achieve high accuracy and low computational costs in time series forecasting is to develop and use efficient hybrid methods. Among the combined methods, parallel hybrid approaches are more welcomed by scholars and often have better performance than sequence ones. However, the necessary condition of using parallel combinational approaches is to estimate the appropriate weight of components. This weighting stage of parallel hybrid models is the most effective factor in forecasting accuracy as well as computational costs. In the literature, meta-heuristic algorithms have often been applied to weight components of parallel hybrid models. However, such that algorithms, despite all unique advantages, have two serious disadvantages of local optima and iterative time-consuming optimization processes. The purpose of this paper is to develop a linear optimal weighting estimator (LOWE) algorithm for finding the desired weight of components in the global non-iterative universal manner.

In this paper, a LOWE algorithm is developed to find the desired weight of components in the global non-iterative universal manner.

Empirical results indicate that the accuracy of the LOWE-based parallel hybrid model is significantly better than meta-heuristic and simple average (SA) based models. The proposed weighting approach can improve 13/96%, 11/64%, 9/35%, 25/05% the performance of the differential evolution (DE), genetic algorithm (GA), particle swarm optimization (PSO) and SA-based parallel hybrid models in electricity load forecasting. While, its computational costs are considerably lower than GA, PSO and DE-based parallel hybrid models. Therefore, it can be considered as an appropriate and effective alternative weighing technique for efficient parallel hybridization for time series forecasting.

In this paper, a LOWE algorithm is developed to find the desired weight of components in the global non-iterative universal manner. Although it can be generally demonstrated that the performance of the proposed weighting technique will not be worse than the meta-heuristic algorithm, its performance is also practically evaluated in real-world data sets.

This study aims to propose iStage, i.e. an intelligent hybrid deep learning (DL)-based framework to determine the stage of the disaster to make the right decisions at the right time.

iStage acquires data from the Twitter platform and identifies the social media message as pre, during, post-disaster or irrelevant. To demonstrate the effectiveness of iStage, it is applied on cyclonic and COVID-19 disasters. The considered disaster data sets are cyclone Fani, cyclone Titli, cyclone Amphan, cyclone Nisarga and COVID-19.

The experimental results demonstrate that the iStage outperforms Long Short-Term Memory Network and Convolutional Neural Network models. The proposed approach returns the best possible solution among existing research studies considering different evaluation metrics – accuracy, precision, recall, f-score, the area under receiver operating characteristic curve and the area under precision-recall curve.

iStage is built using the hybrid architecture of DL models. It is effective in decision-making. The research study helps coordinate disaster activities in a more targeted and timely manner.

Aiming at the problems of weak generalization of robot imitation learning methods and higher accuracy requirements of low-level detectors, this study aims to propose an imitation learning method based on structural grammar.

The paper proposes a hybrid training model based on artificial immune algorithm and the Baum–Welch algorithm to extract the action information of the demonstration activity to form the {action-object} sequence and extract the symbol description of the scene to form the symbol primitives sequence. Then, probabilistic context-free grammar is used to characterize and manipulate these sequences to form a grammar space. Minimum description length criteria are used to evaluate the quality of the grammar in the grammar space, and the improved beam search algorithm is used to find the optimal grammar.

It is found that the obtained general structure can parse the symbol primitive sequence containing noise and obtain the correct sequence, thereby guiding the robot to perform more complex and higher-order demonstration tasks.

Using this strategy, the robot completes the fourth-order Hanoi tower task has been verified.

An imitation learning method for robots based on structural grammar is first proposed. The experimental results show that the method has strong generalization ability and good anti-interference performance.

This chapter presents a combined approach of social and open data to evaluate a hybrid education model with online and face-to-face classes. The study consists of a sample of 310 students from the UPIITA-IPN college. Thus, a grouping model was applied based on each student's profile and academic performance in various subjects to identify patterns and learning styles. In addition, a social sensor of emotions was implemented to measure reactions in online and face-to-face classes. It helped to identify which strategies and methods are most significant for student performance. Data were collected from forms and the Twitter social network, filtering data by general opinions about learning and experiences in class. Considering trends and patterns, we identified four types:

Pattern (1) personalization of learning: This group stood out because online teaching allows you to work at your own pace and on your own schedule. In addition, a trend toward a more individualized learning approach or the versatility of personalizing learning was observed. Pattern (2) an excessive number of channels and information: This group of students was characterized by feeling overwhelmed by the amount of information they must process in an online environment, in addition, to using various communication channels (messaging, Classroom, Zoom, Teams, email, among others) this was associated with a feeling of isolation and a lack of commitment. Pattern (3) inequality and asynchronous learning: Students with difficult access to adequate resources at home (connection, own computer, etc.). They were characterized by not being able to have the same performance in the different learning activities and expressed that the content must be adapted to the individual needs of the students. Technical problems, such as Internet connection failures or problems with electronic devices, interrupted the learning process and generated frustration for students and teachers. Pattern (4) lack of social interaction: This affected the student's ability to develop social and emotional skills. Moreover, it generates difficulties for the students to collaborate, slowing the development of social and emotional skills. It concluded that a hybrid model is successful, having schemes combined with 65% face-to-face sessions and 35% online.

In community question answering (CQA), people who answer questions assume readers have mastered the content in the answers. Nevertheless, some readers cannot understand all content. Thus, there is a need for further explanation of the concepts that appear in the answers. Moreover, the large number of question and answer (Q&A) documents make manual retrieval difficult. This paper aims to alleviate these issues for CQA websites.

In the paper, an algorithm for recommending explanatory Q&A documents is proposed. Q&A documents are modeled with the biterm topic model (BTM) (Yan et al., 2013). Then, the growing neural gas (GNG) algorithm (Fritzke, 1995) is used to cluster Q&A documents. To train multiple classifiers, three features are extracted from the Q&A categories. Thereafter, an ensemble classification model is constructed to identify the explanatory relationships. Finally, the explanatory Q&A documents are recommended.

The GNG algorithm shows good clustering performance. The ensemble classification model performs better than other classifiers. The both effect and quality scores of explanatory Q&A recommendations are high. These scores indicate the practicality and good performance of the proposed recommendation algorithm.

The proposed algorithm alleviates information overload in CQA from the new perspective of recommending explanatory knowledge. It provides new insight into research on recommendations in CQA. Moreover, in practice, CQA websites can use it to help retrieve Q&A documents and facilitate understanding of their contents. However, the algorithm is for the general recommendation of Q&A documents which does not consider individual personalized characteristics. In future work, personalized recommendations will be evaluated.

A novel explanatory Q&A recommendation algorithm is proposed for CQA to alleviate the burden of manual retrieval and Q&A overload. The novel GNG clustering algorithm and ensemble classification model provide a more accurate way to identify explanatory Q&A documents. The method of ranking the explanatory Q&A documents improves the effectiveness and quality of the recommendation. The proposed algorithm improves the accuracy and efficiency of retrieving explanatory Q&A documents. It assists users in grasping answers easily.