MACHINE LEARNING MODELS: APPLICATION AND EFFICIENCY

Abstract

There are many systems that are studied and developed in the field of machine learning and artificial intelligence which are constantly evolving fields that have transformed a multitude of industries and applications around the world. The purpose of this study lies in a comprehensive review of 120 articles, highlighting the diversity of data types used in machine learning, from structured data such as tables and time series to unstructured data such as images and text. To develop this research, a systematic review of the literature was carried out, originating from the search for computational models used in different areas of human knowledge; The search proceeded with the execution of the data extraction, describing each of the fields considered for the analysis of the information; The data was processed and analyzed considering the models with the highest frequency of use and their performance metrics. The results of this research highlight the preference of the Support Vector Machines model as the most frequently used in a variety of applications. Likewise, the research also reveals that, in terms of efficiency and precision, the Gradient Boosting and Artificial Neural Networks models stand out significantly. These results evidence the importance of interdisciplinary collaboration and the need for proper application to ensure that machine learning and artificial intelligence continue to be drivers for technological advancement.