Search

The state of Maranhao has low levels of quality in basic education, as can be seen in the national performance assessments over the years. The problem of low educational quality can be approached and investigated from the view point of several areas. One of them is the use of Educational Data Mining, which is increasingly present in scientific studies, and is also used to support decision-making by public policy makers. However, research with data from a geographic region, or a particular location, may still be scarce, as is the case in the state of Maranhao. This research aims to understand what are the factors that influence the quality of public education in the state of Maranhao. For this purpose, data mining techniques were used, such as exploratory data analysis, correlation analysis, factor analysis, regression and decision tree. The data used are from state public high schools in Maranhao. Results of this study show a diagnosis of the state's educational situation, with the identification of what significantly influences the performance of education. [ABSTRACT FROM AUTHOR]

Analyzing the relation between socioeconomic variables and neglected tropical diseases can help managers in the conception of public policies to reduce cases. The objective of this study was to evaluate, based on machine learning algorithms, which socioeconomic variables are more important for the risk classification of three neglected diseases: leprosy, cutaneous leishmaniasis, and dengue. Three algorithms based on decision trees were evaluated: Random Forest (RF), XGBoost, and C5.0. As a study area, the municipalities of the state of Goiás and of the Federal District - Brazil, were delimited. For the dengue risk classes, both the RF algorithm and the XGBoost showed accuracy values above 0.6. Both emphasizing the low-income conditions, literacy, and race as the most important predictive variables. In the leprosy risk classes case, the three algorithms presented accuracy results above 0.6, indicating the variables water supply, literacy, race, and housing as important. For the tegumentary leishmaniasis risk classes, the algorithms showed an accuracy lower than 0.4, making the evaluation of possible predictive variables to the model unfeasible. The three evaluated algorithms revealed approximate predictive performance; however, the RF was slightly higher. The most important socioeconomic variables for dengue and leprosy risk classes prediction were similar. [ABSTRACT FROM AUTHOR]

The growth of digital documents, associated with their usage in several knowledge areas requires computational resources for its comprehension and analysis. The literature proposes distinguishing authors by their writing style and keywords. However, these studies mainly involve journalistic and literary contexts written in English. This research is unique because it explores authorship analysis within a dataset composed of school activities written by undergraduate students in Portuguese. Such a scenario is challenging because it contains fewer documents per author, homogeneous authors, and fewer research and tools in Portuguese. Due to the insufficient number of samples, we used robust journalistic datasets as reference. The experiments verified that stylometric representations are superior to textual representations in restricted domains, which suffer from the topic's broader corpora. Furthermore, we found out that the ensemble of extremelly randomized decision trees associated with the proposed stylometric features overcome every other model tested, in allthe datasets, reaching an average accuracy of 0.71 and 0.81 AUC. [ABSTRACT FROM AUTHOR]

The relief is in constant transformation due to natural agents and processes. The Digital Terrain Model (MDT) emerges as an important tool in obtaining the altimetric information of the relief, enabling an automated analysis that allow interpolations of pre-defined variables. The present study aims at geomorphometric compartmentation of the Jaguari River Hydrographic Basin from the decision tree and interpolation of data related to relief morphometry, by correlating the topographic data (altitude, slope, curvature profile and plane of curvature) generated in a GIS environment. Eighteen (18) geomorphometric units were defined, which enabled the analysis of the flow behavior and erosive processes in the slopes. [ABSTRACT FROM AUTHOR]

Sepsis is a generalized inflammation with high morbidity and mortality, whose recognition and early treatment are essential factors for a better patient's quality of life; if not identified and treated promptly, could lead to death. This integrative review article aims to identify the techniques based on artificial intelligence adopted, their respective accuracy, sensitivity and specificity for early identification in cases of sepsis in a hospital environment. The research, adapted from the PRISMA method, was performed in five databases indexed from the following descriptors: sepsis, septic, forecasting, predict, prediction, detection, predicting, diagnosis, assessment, machine learning, artificial intelligence, data mining and deep learning. A total of 333 articles were identified, 21 with reference to the early recognition of sepsis by 16 techniques. The results showed that the neural networks performed better, varying the accuracy between 76% and 93%, the decision trees between 69.0% and 91.5% and the statistical methods between 56% and 89%. It is concluded that the most influential factor in the early identification of the diagnosis is the variety and quality of the data. The challenge in relation to pre-processing is also evident, since the data are generally from heterogeneous sources, collected with different criteria, methods and objectives. [ABSTRACT FROM AUTHOR]

This article deals with the value of information in a natural gas exploration and production project. The financial analysis of natural gas fields is a fundamental part in the decision making process of investments by companies of the sector, government and agents of the financial market. An effective analysis can save important economic resources or even generate them when everything else indicates the opposite. During the stages of the exploration and production project, the concessionaire makes high investments in the acquisition of geological data, which increase the knowledge level of the asset in question. This case study examines this issue in light of the Real Options Theory, using the Bayesian inversion method to quantify the value of the information, exploring abandon and expansion real options. The relevance of this article is the presentation of an evaluation model in its state-of-the-art applied in a segment of high importance to the Brazilian economy. [ABSTRACT FROM AUTHOR]

Land cover is a spatial information of great relevance for a variety of models for estimating sediment yield and to measure the potential of the landscape carbon sequestration. The classification of land cover by the supervised method requires training areas, these areas must be representative of each class of land cover. For the classification decision tree (DT) algorithm, the complexity of DT, results in different values of accuracies for thematic maps. Thus, the objective of this study was to estimate the minimum sample density in a DT model which would allow to discriminate land cover classes, evaluate the size of the generated DT model, as well as, identify the more difficult land cover class to be mapped. Satellite images from RESOURCESAT-1 as well as spectral indices were used in the study. The minimum sample density varied between 0.15 and 0.30% of the total area for each class, this sampling interval allowed better results than 80% for kappa index. The smallest grouping of observations in the same terminal leaf was 45 observations. In this study the most difficult land use classes to be mapped were forest and rice crops due to spectral similarity of shaded forests with irrigated rice crops. [ABSTRACT FROM AUTHOR]

One of the challenges of precision agriculture is to offer subsidies for the definition of management units for posterior interventions. Therefore, the objective of this work was to evaluate soil chemical attributes and sugarcane yield with the use of geostatistics and data mining by decision tree induction. Sugarcane yield was mapped in a 23ha field, applying the cell criterion, by using a yield monitor that allowed the elaboration of a digital map representing the surface of production of the studied area. To determine the soil attributes, soil samples were collected at the beginning of the harvest in 2006/2007 using a regular grid of 50 x 50m, in the depths of 0.0-0.2m and 0.2-0.4m. Soil attributes and sugarcane yield data were analyzed by using geostatistics techniques and were classified into three yield levels for the elaboration of the decision tree. The decision tree was induced in the software SAS Enterprise Miner, using an algorithm based on entropy reduction. Altitude and potassium presented the highest values of correlation with sugarcane yield. The induction of decision trees showed that the altitude is the variable with the greatest potential to interpret the sugarcane yield maps, then assisting in precision agriculture and, revealing an adjusted tool for the study of management definition zones in area cropped with sugarcane. [ABSTRACT FROM AUTHOR]