Sabina-Adriana Floria

This paper proposes a classification methodology aimed at identifying correlations between job ad requirements and transversal skill sets, with a focus on predicting the necessary skills for individual job descriptions using a deep learning model. The approach involves data collection, preprocessing, and labeling using ESCO (European Skills, Competences, and Occupations) taxonomy. Hierarchical classification and multi-label strategies are used for skill identification, while augmentation techniques address data imbalance, enhancing model robustness. A comparison between results obtained with English-specific and multi-language sentence embedding models reveals close accuracy. The experimental case studies detail neural network configurations, hyperparameters, and cross-validation results, highlighting the efficacy of the hierarchical approach and the suitability of the multi-language model for the diverse European job market. Thus, a new approach is proposed for the hierarchical classification of transversal skills from job ads.

Artificial neural networks have proven to be effective in a wide range of fields, providing solutions to various problems. Training artificial neural networks using evolutionary algorithms is known as neuroevolution. The idea of finding not only the optimal weights and biases of a neural network but also its architecture has drawn the attention of many researchers. In this paper, we use different biologically inspired optimization algorithms to train multilayer perceptron neural networks for generating regression models. Specifically, our contribution involves analyzing and finding a strategy for combining several algorithms into a hybrid ensemble optimizer, which we apply for the optimization of a fully connected neural network. The goal is to obtain good regression models for studying and making predictions for the process of free radical polymerization of methyl methacrylate (MMA). In the first step, we use a search procedure to find the best parameter values for seven biologically inspired optimization algorithms. In the second step, we use a subset of the best-performing algorithms and improve the search capability by combining the chosen algorithms into an ensemble of optimizers. We propose three ensemble strategies that do not involve changes in the logic of optimization algorithms: hybrid cascade, hybrid single elite solution, and hybrid multiple elite solutions. The proposed strategies inherit the advantages of each individual optimizer and have faster convergence at a computational effort very similar to an individual optimizer. Our experimental results show that the hybrid multiple elite strategy ultimately produces neural networks which constitute the most dependable regression models for the aforementioned process.

Neural networks have demonstrated their usefulness for solving complex regression problems in circumstances where alternative methods do not provide satisfactory results. Finding a good neural network model is a time-consuming task that involves searching through a complex multidimensional hyperparameter and weight space in order to find the values that provide optimal convergence. We propose a novel neural network optimizer that leverages the advantages of both an improved evolutionary competitive algorithm and gradient-based backpropagation. The method consists of a modified, hybrid variant of the Imperialist Competitive Algorithm (ICA). We analyze multiple strategies for initialization, assimilation, revolution, and competition, in order to find the combination of ICA steps that provides optimal convergence and enhance the algorithm by incorporating a backpropagation step in the ICA loop, which, together with a self-adaptive hyperparameter adjustment strategy, significantly improves on the original algorithm. The resulting hybrid method is used to optimize a neural network to solve a complex problem in the field of chemical engineering: the synthesis and swelling behavior of the semi- and interpenetrated multicomponent crosslinked structures of hydrogels, with the goal of predicting the yield in a crosslinked polymer and the swelling degree based on several reaction-related input parameters. We show that our approach has better performance than other biologically inspired optimization algorithms and generates regression models capable of making predictions that are better correlated with the desired outputs.

In the brick manufacturing industry, there is a growing concern among researchers to find solutions to reduce energy consumption. An industrial process for obtaining bricks was approached, with the manufacturing mix modified via the introduction of sunflower seed husks and sawdust. The process was analyzed with artificial intelligence tools, with the goal of minimizing the exhaust emissions of CO and CH₄. Optimization algorithms inspired by human and virus behaviors were applied in this approach, which were associated with neural network models. A series of feed-forward neural networks have been developed, with 6 inputs corresponding to the working conditions, one or two intermediate layers and one output (CO or CH₄, respectively). The results for ten biologically inspired algorithms and a search grid method were compared successfully within a single objective optimization procedure. It was established that by introducing 1.9% sunflower seed husks and 0.8% sawdust in the brick manufacturing mix, a minimum quantity of CH₄ emissions was obtained, while 0% sunflower seed husks and 0.5% sawdust were the minimum quantities for CO emissions.

Online social networks are the main choice of people to maintain their social relationships and share information or opinions. Estimating the actions of a user is not trivial because an individual can act spontaneously or be influenced by external factors. In this paper we propose a novel model for imitating the evolution of the information diffusion in a network as well as possible. Each individual is modeled as a node with two factors (psychological and sociological) that control its probabilistic transmission of information. The psychological factor refers to the node’s preference for the topic discussed, i.e. the information diffused. The sociological factor takes into account the influence of the neighbors’ activity on the node, i.e. the gregarious behavior. A genetic algorithm is used to automatically tune the parameters of the model in order to fit the evolution of information diffusion observed in two real-world datasets with three topics. The reproduced diffusions show that the proposed model imitates the real diffusions very well.

Social networks currently belong to a vast area of research as information spreads at a remarkable speed due to technology, and social connections have become easily accessible in the online environment. Social networks are dynamic entities, which new individuals can join, or other links can be lost because members no longer interact with one-another. Dynamic analysis of social networks is important in topology changes of the network and also in information diffusion. Some information that spreads through the social network may be untrue, hence in this paper we propose a protocol based on evidence theory with Dempster-Shafer and Yager’s rules in which the network becomes more immune to false information. We also analyze the impact of topology change for an initial network by adding new connections in the information diffusion process. We show information diffusion by coloring the nodes of the network and also illustrate the time evolution of messages for a better accuracy in our comparisons. The experimental results confirm that the proposed model fits the behavior of inhibiting false information.

This paper deals with the complex problem of reliability evaluation of stochastic networks in which both links and nodes failures are considered, and compares two approximate approaches able to reduce the computation time: a sum of disjoint products (SDP) approach and another based on Monte Carlo simulation. In case of the SDP approach, the reliability is computed based on the minimal paths. In a first stage, only the links of the network are considered, and a `multiple variables inversion' technique for developing the set of minimal paths into a sum of disjoint products is applied. Then, in a second stage, each term of the set of disjoint Iink- products is processed separately taking into consideration the reliability values for both links and adjacent nodes. In case of Monte Carlo simulation, for speeding up the method and reducing the computation time, both minimal paths and cuts are considered. Also, other acceleration techniques are applied.