PERFORMANCE ANALYSIS OF REGRESSION ALGORITHMS USING SCIKIT-LEARN.

Abstract

Machine learning and deep learning are currently áreas of extremely active research, since they allow projections and estimates of variables based on historical data and the execution of mathematical algorithms; These algorithms are specialized to performregressions, classification, prediction or other applications where they are used large volumes of data and decisions with a high margin of precision are required [1]. The large amount of data available today provides great opportunities and the transformative potential of different sectors such as medicine, banking, engineering, sports, among others; but they also present great challenges in the use information, since a poor or erroneous analysis of the data leads to to wrong decision making. As data grows, learning deep takes a leading role in the analysis and solution of problems with a high degree of complexity that in a natural environment are not easy to solve because they present nuances that only with the use of deep learning algorithms can be observed. [two]. Improvements in computational power, large volumes of information, fast data storage and parallelization have contributed to the analysis and prediction of Big Data in areas such as price prediction, análisis of medical images, traffic control and even the study of the performance of the football team, among many others. With all of the above, an idea of ​​the current importance of this area of ​​science is given and how pertinent it is to work on this topic