Systems and computers engineering

In this paper, hierarchical control architecture for coordinated multi-robot systems (MRS) task decomposition is presented; based on a hybrid decentralized Partially Observable Semi-Markov Decision Processes (HDec-POSMDPs).  In this architecture, robots can make their own decisions according to their locally collected data with limited communication between a robot team.  In this proposed architecture, the global task is decomposed into multiple local sub-tasks using divide and conquer design, each task is described as a set of regular languages. MRS are modeled as a discrete event system and each robot is represented by a deterministic finite state automaton model. Direct Cross-Entropy (DICE) can be used for searching the space of the best frontier cells to solve the Dec-POSMDP and each sub-task is assigned to one or more robots to be executed. The proposed algorithm is implemented, tested and evaluated in the computer simulator. By using this architecture, the task execution time ...

Nowadays, due to the benefits of the agile method, such as changing rapidly and fast delivery of working software, most software projects are naturally distributed agile projects (DAD). DAD team members work from various remote sites. This leads to the emergence of many significant challenges in risk management. Hence, it requires risk safety techniques to be implemented to mitigate the occurrence of risks. There is no standardized process for teams to deal with DAD risks. In this paper, a multi-agent reinforcement learning risk management model for distributed agile software projects is proposed. The proposed model is implemented to apply a dynamic policy. The model is applied as an experiment to definite numbers for a set of risk factors such as communication and coordination risk, project management risk, and SDLC risk. The proposed model is developed using multiple individual learning using the Q- learning algorithm. A DAD project with two projects is used to evaluate the proposed model. The proposed model was assessed and analyzed for its effectiveness. The results of the evaluation are given.

Some of the educational organizations have multi-education paths such as engineering and medicine collages. In such colleges, the behavior of the student in the preparatory year determines which education path the student will join in the future. In this paper, an adaptive recommendation system is proposed for predicting a suitable education path(s) for a student in college preparatory year. The adaptability is achieved by automatically applying different data mining techniques for extracting relevant features and building a tailor-made model for each education path. The problem formulated as a multi-label multi-class binary classification problem and the dataset automatically translated into one-versus-all (for binary classification). As a case study, the proposed model is applied to predict student’s academic performance in the faculty of engineering at AL-Azhar University. It recommends a suitable engineering department among seven engineering departments for each student based on his academic performance. The data of each department (i.e. educational program) is fed to the recommendation system. Then, the relevant set of features for each department is selected and a machine learning algorithm with the best performance is selected for the recommendation process of each department. The obtained results showed that the proposed model recommends the best machine learning algorithm (i.e. model) for each faculty department, find the relevant data that are important in the recommendation process and recommend the student with the suitable engineering department(s) with high accuracy.

Recently, a lot of Biometric Authentication Systems (BAS) techniques for the purpose of security in both industrial and academic organizations because it improves the authentication system's usability by overcoming the existing difficulties of using passwords. Although of the rapid spread of BAS’s systems especially in smartphones, it still has defects and are subjected to attacks, and suffers from the intrusion of user privacy, because some biometric features have not been deeply studied and investigated. In this survey, the existing BAS techniques are explored and classified focusing on the security issues that they are used for and the corresponding solutions provided in this domain. The advantages and disadvantages of each BAS technique are also reviewed, the threats of each BAS technique are listed, summarized, and discussed. The relevant research studies are also presented and several open problems, research direction are also presented.

Model Predictive Control (MPC) has been very successful in the control industries due to its ability to handle uncertainty in physical systems with some constraints. Due to its higher power density, DC servomechanism is used in control systems of several industrial applications. However, the DC servomechanisms suffer severe throttling loss, leading to low effectiveness, and the uncertainty parameters and distortions caused by the changing system environment harm their performance. In this paper, a Robust Integral MPC (RIMPC) model is designed to enhance the efficiency of the DC servomechanism control system. Uncertainties and nonlinearity of the parameters are taken into consideration. In RIMPC, MPC is incorporated into the integral controller synthesis through an optimization problem subjected to some Linear Matrix Inequality (LMI) constraints to generate the control law of the servomechanism dynamics. The performance of the proposed system is evaluated in two cases: with no disturbances and when disturbances are affecting the system. The obtained results are analyzed and compared with other controllers' design, such as Proportional Integral Derivative (PID) controller and the Monte Carlo (MC) controller. The obtained results prove that the proposed system is more effective in cost function and transient performance (rise time, peak time, maximum overshoot, and settling-time), indicating a much better reference tracking accuracy and proposed method robustness.

Environment Exploration is the basic process that most of Multi Robot Systems applications depend on it. The exploration process performance depends on the coordination strategy between the robots participating in the team.  In this paper the coordination of Multi Robot Systems in the exploration process is surveyed, and the performance of different Multi Robot Systems exploration strategies is contrasted and analyzed for different environments and different team sizes.

Exploring the environment using multi-robot systems is a fundamental process that most automated applications depend on. This paper presents a hybrid decentralized task assignment approach based on Partially Observable Semi-Markov Decision Processes called HDec-POSMDPs, which are general models for multi-robot coordination and exploration problems in which robots can make their own decisions according to its local data with limited communication between the robot team. In this paper, a variety of multi-robot exploration algorithms and their comparison have been tackled. These algorithms, which have been taken into consideration, are dependent on different parameters. Collectively, there are five metrics maximize the total exploration percentage, minimize overall mission time, reduce the number of hops in the networked robots, reduce the energy consumed by each robot and minimize the number of turns in the path from the start pose cells to the target cells. Therefore, a team of ident...

Dialog Generation Systems (DGS) have emerged as a critical aspect of Natural Language Processing in recent years (NLP). It enables a diverse set of relevant applications to interact with humans in a natural and intelligent way. This study provides a systematic review of open DGS techniques that have recently been used. The major goal of this study is to discuss and analyze the most widely used approaches for implementing DGS's that have been published in recent years. Also, the most popular datasets for open DGS are enumerated, and some commonly used automatic evaluating metrics are presented. As a result, the explored methods are categorized into six main categories, Reinforcement Learning (RL), Hierarchical Recurrent Encoder-Decoder (HRED), Generative Adversarial Networks (GAN), Variational Auto-Encoder (VAE), Sequence to Sequence (Seq2Seq), and Pretraining Model.