Bringing Machine Learning Predictive Models Based on Machine Learning Closer to Non-technical Users
| Autor: | Pablo Pico-Valencia, Juan A. Holgado-Terriza, Oscar Vinueza-Celi |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
business.industry
Computer science Computer programming Supervised learning 020206 networking & telecommunications 02 engineering and technology Python (programming language) Machine learning computer.software_genre Software Scripting language 0202 electrical engineering electronic engineering information engineering Unsupervised learning 020201 artificial intelligence & image processing Artificial intelligence business MATLAB computer computer.programming_language |
| Zdroj: | Systems and Information Sciences ISBN: 9783030591939 |
| DOI: | 10.1007/978-3-030-59194-6_1 |
| Popis: | Today, data science has positioned as an area of interest for decision makers in many organizations. Advances in Machine Learning (ML) allow training predictive models based on the analysis of datasets in multiple domains such as: business, medicine, marketing, among others. These models are able to learn and predict future behaviors which helps in the decision-making process. However, many of the ML tools such as Python, Matlab, R Suite, and even their libraries, require that every action must be performed as a sequence of commands by means of scripts. These software packages require extensive technical knowledge of statistics, artificial intelligence, algorithms and computer programming that generally only computer engineers are skilled at. In this research we propose the development of a process complemented with the assistance of a set of user graphic interfaces (GUIs) to help non-sophisticated users to train and test ML models without writing scripts. A tool compatible with Python and Matlab was developed with a set of GUIs adapted to professionals of the business area that generally require to apply ML models in their jobs, but they do not have time to learn programming. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|