Generally, decision making in urban planning has progressively become difficult due to the uncertain, convoluted, and multi-criteria nature of urban issues. Even though there has been a growing interest to this domain, traditional decision support systems are no longer able to effectively support the decision process. This paper aims to elaborate an intelligent decision support system (IDSS) that provides relevant assistance to urban planners in urban projects. This research addresses the use of new techniques that contribute to intelligent decision making: machine learning classifiers, naïve Bayes classifier, and agglomerative clustering. Finally, a prototype is being developed to concretize the proposition.

No abstract

<p> • Background Enhancing the resiliency of electric power grids is becoming a compelling issue, especially when considering the acute outages that occurred recently in different parts of the world. One preeminent improvement of those grids should be at the level of anticipating the imminent failure that may be engendered by line contingency or grid disturbances. This being the case, a number of researchers in the field of power industry have initiated some investigations to generate some techniques of predicting power outages. Still, extended blackouts can occur due to the vulnerability of the distribution power grid in case of extreme events. </p><p> • Objective This paper implements a proactive prediction model based on deep-belief networks that can predict imminent blackouts departing from previous historical blackouts, trigger alarms and suggest solutions in case those blackouts take place. These actions can prevent outages, stop cascading failures and diminish possible economic losses caused by power outages. </p><p> • Method The proposed deep learning based model is divided into three phases: A, B and C. The first phase A represents the initial segment that is used to collect and extract the necessary data and train the deep belief network later, using the collected data. Phase B defines the Power outage threshold, and decides whether the grid is in a normal state or not. Phase C involves detecting supplementary dangerous events, triggering alarms and proposing an emergency actions plan that helps in the process of power restoration. </p><p> • Results Different machine learning and deep learning algorithms are used in our experiments such as Random forest, Bayesian nets and others in order to validate our proposition. Deep belief Networks offer a rate of 97.30% as accuracy and 97.06% as precision rate. </p><p> • Conclusion According to the research that has been carried out, it is possible to conclude that our model could be used perfectly for blackouts’ prediction and that the deep-belief network represents a powerful deep learning tool that could offer quite plausible results.

No abstract

No abstract