The recurrent use of databases with categorical variables in different applications demands new alternatives to identify relevant patterns. Classification is an interesting approach for the recognition of this type of data. However, there are a few amount of methods for this purpose in the literature. Also, those techniques are specifically focused only on kernels, having accuracy problems and high computational cost. For this reason, we propose an identification approach for categorical variables using conventional classifiers (LDC-QDC-KNN-SVM) and different mapping techniques to increase the separability of classes. Specifically, we map the initial features (categorical attributes) to another space, using the Chi-square (C-S) as a measure of dissimilarity. Then, we employ the (t-SNE) for reducing dimensionality of data to two or three features, allowing a significant reduction of computational times in learning methods. We evaluate the performance of proposed approach in terms of accuracy for several experimental configurations and public categorical datasets downloaded from the UCI repository, and we compare with relevant state of the art methods. Results show that C-S mapping and t-SNE considerably diminish the computational times in recognitions tasks, while the accuracy is preserved. Also, when we apply only the C-S mapping to the datasets, the separability of classes is enhanced, thus, the performance of learning algorithms is clearly increased.

The education is a fundamental law that promotes the social and economic development of a nation. Also, it is how a country can achieve its sustainable development objectives. Because of this, various international organizations (Unicef-UNESCO- Banco Mundial -OECD) are promoting the coverage of education in underdeveloped countries. On the other hand, projects such as PISA that aim to carry out annual evaluations in more than 30 countries that serve as a reference of the educational level and encourages nations to improve the quality of education. Therefore, the objective of this work is to implement a methodology to predict the results of ICFES tests SABER 11 °. that can correct the educational problems that Colombia has presented in secondary education. For the development of this methodology, the ICFES repository database was used. Data were pre-processed using MATLAB software. Also, combined tests were carried out with the new Chi-Square metric developed by the researchers of the automation group of the Technological University of Pereira, obtaining results that were 20% higher compared with conventional classification techniques. In this project, the most influential characteristics were found in the students, who are responsible for the loss of the ICFES SABER 11° exam in Risaralda.

Este trabajo de investigación presenta el desarrollo de un modelo utilizando técnicas de minería de datos para identificar variables financieras en una empresa manufacturera de carrocerías para vehículos automotores en Pereira. El estudio se estructura en cuatro fases claves. La primera fase se centra en el preprocesamiento de datos, incluyendo la caracterización, normalización, y reducción de dimensionalidad mediante PCA, Relief y Correlación. La segunda fase aplica aprendizaje no supervisado con K-means y Mezclas Gaussianas (GMM) para agrupar y validar datos según una variable objetivo definida. En la tercera fase, se emplean clasificadores supervisados como el Clasificador Bayesiano, Redes Neuronales Artificiales, Máquinas de Soporte Vectorial y KNN para predecir la eficiencia de los proveedores, optimizando los procesos de inversión y costeo. Finalmente, la cuarta fase integra el preprocesamiento y la predicción en un formulario práctico, utilizando librerías como Plotly y Dash para visualizaciones detalladas, y herramientas como GitHub y Heroku para el desarrollo de la aplicación. Este estudio destaca la importancia de la inteligencia artificial en la toma de decisiones empresariales, demostrando cómo las técnicas de ciencia de datos y las herramientas de visualización pueden facilitar la interpretación y el aprovechamiento de los resultados del análisis de datos.