Electricity distribution companies have been incorporating new technologies that allow them to obtain complete information in real time about their customers´ consumption. Thus, a new concept called ..."Smart Metering" has been adopted, giving way to new types of meters that interact in an interconnected system. This will allow to make data analysis, accurate forecasts and detecting consumption patterns that will be relevant for the decision-making process. This research focuses on discovering common patterns among customers from data collected by smart meters.
In higher education, student dropout is a relevant problem, not just in Latin America but also in developed countries. Although there is no consensus to measure the education quality, one of the ...important indicators of university success is the time to graduation (TTG), which is directly related to student dropout 1. Global estimates put this dropout rate at 42% 2. In the United States, this rate is around 30% and represents a loss of 9 billion dollars in the education of these students 3. However, desertion not only affects the quality of education and the economy of a country, but also has effects on the development of society, since society demands the contributions derived from the population with higher education such as: innovation, knowledge production and scientific discovery 4. Using basic statistical learning techniques, this paper presents a simple way to predict possible dropouts based on their demographic and academic characteristics.
Data Mining Applied in School Dropout Prediction Viloria, Amelec; García Guliany, Jesús; Niebles Núñez, William ...
Journal of physics. Conference series,
01/2020, Letnik:
1432, Številka:
1
Journal Article
Recenzirano
Odprti dostop
In recent years, many studies have emerged about regarding the topic of school failure, showing a growing interest in determining the multiple factors that may influence it 1. Most of the researches ...that attempt to solve this issue 2 are focused on determining the factors that most affect the performance of students (dropout and failure) at the different educational levels (basic, middle and higher education) through the use of the large amount of information that current computer equipment allows to store in databases. All these data constitute a real gold mine of valuable information about students. But, identifying and finding useful and hidden information in large databases is a difficult task 3. A very promising solution to achieve this goal is the use of knowledge mining techniques or data mining in education, which has resulted in so-called Educational Data Mining (EDM) 4. This new area of research is concerned with the development of methods for exploring data in education, as well as the use of these methods to better understand students and the contexts where they learn 5.
In the new global and local scenario, the advent of intelligent distribution networks or Smart Grids allows real-time collection of data on the operating status of the electricity grid. Based on this ...availability of data, it is feasible and convenient to predict consumption in the short term, from a few hours to a week. The hypothesis of the study is that the method used to present time variables to a prediction system of electricity consumption affects the results.
Neural Networks for Tea Leaf Classification Silva, Jesús; Hernández Palma, Hugo; Niebles Núñez, William ...
Journal of physics. Conference series,
01/2020, Letnik:
1432, Številka:
1
Journal Article
Recenzirano
Odprti dostop
The process of classification of the raw material, is one of the most important procedures in any tea dryer, being responsible for ensuring a good quality of the final product. Currently, this ...process in most tea processing companies is usually handled by an expert, who performs the work manually and at his own discretion, which has a number of associated drawbacks. In this work, a solution is proposed that includes the planting, design, development and testing of a prototype that is able to correctly classify photographs corresponding to samples of raw material arrived at a dryer, using intelligence techniques (IA) type supervised for Classification by Artificial Neural Networks and not supervised with K-means Grouping for class preparation. The prototype performed well and is a reliable tool for classifying the raw material slammed into tea dryers.
This study describes a model of explanations in natural language for classification decision trees. The explanations include global aspects of the classifier and local aspects of the classification ...of a particular instance. The proposal is implemented in the ExpliClas open source Web service 1, which in its current version operates on trees built with Weka and data sets with numerical attributes. The feasibility of the proposal is illustrated with two example cases, where the detailed explanation of the respective classification trees is shown.
Data Mining is the process of analyzing data using automated methodologies to find hidden patterns 1. Data mining processes aim at the use of the dataset generated by a process or business in order ...to obtain information that supports decision making at executive levels 2 3 through the automation of the process of finding predictable information in large databases and answer to questions that traditionally required intense manual analysis 4. Due to its definition, data mining is applicable to educational processes, and an example of that is the emergence of a research branch named Educational Data Mining, in which patterns and prediction search techniques are used to find information that contributes to improving educational quality 5. This paper presents a performance study of data mining algorithms: Decision Tree and Logistic Regression, applied to data generated by the academic function at a higher education institution.
This paper proposes an innovative way to address real cases of production prediction. This approach consists in the decomposition of original time series into time sub-series according to a group of ...factors in order to generate a predictive model from the partial predictive models of the sub-series. The adjustment of the models is carried out by means of a set of statistic techniques and Automatic Learning. This method was compared to an intuitive method consisting of a direct prediction of time series. The results show that this approach achieves better predictive performance than the direct way, so applying a decomposition method is more appropriate for this problem than non-decomposition. The agricultural sector will be used as the study subject.
This paper proposes the analysis of the influence of terms that express feelings in the automatic detection of topics in social networks. This proposal uses an ontology-based methodology which ...incorporates the ability to identify and eliminate those terms that present a sentimental orientation in social network texts, which can negatively influence the detection of topics. To this end, two resources were used to analyze feelings in order to detect these terms. The proposed system was evaluated with real data sets from the Twitter and Facebook social networks in English and Spanish respectively, demonstrating in both cases the influence of sentimentally oriented terms in the detection of topics in social network texts.
Technological advances have allowed to collect and store large volumes of data over the years. Besides, it is significant that today's applications have high performance and can analyze these large ...datasets effectively. Today, it remains a challenge for data mining to make its algorithms and applications equally efficient in the need of increasing data size and dimensionality 1. To achieve this goal, many applications rely on parallelism, because it is an area that allows the reduction of cost depending on the execution time of the algorithms because it takes advantage of the characteristics of current computer architectures to run several processes concurrently 2. This paper proposes a parallel version of the FuzzyPred algorithm based on the amount of data that can be processed within each of the processing threads, synchronously and independently.