Machine learning for B2B manufacturing price prediction
Maldonado Sada, Daniela (2023)
Diplomityö
2023
School of Engineering Science, Laskennallinen tekniikka
Kaikki oikeudet pidätetään.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe20230829112158
https://urn.fi/URN:NBN:fi-fe20230829112158
Tiivistelmä
The B2B industry encounters pricing challenges that usually require direct market research. However, real-time transactional data analysis can enhance business profitability by utilizing machine learning methods.
The aim is to have an overview of various machine learning algorithms widely used in today's data-driven world. Among the models discussed are decision trees, Bayesian networks, principal component regression, support vector regression, neural networks, linear regression, and Gaussian regression. We will examine each model's unique features to determine which is best for our data.
This thesis analyzes data via visualization and correlation matrix to understand what type of model we need according to our data. However, the primary objective is to assess the practicality and restrictions of each of the preselected models.
A weighted decision matrix will assist us to make the comparative analysis and select the optimal model from the three preselected models (decision trees, Gaussian processes, and linear regressions). The findings of this study may be beneficial for scholars working in the areas of data science, statistics, and machine learning.
The aim is to have an overview of various machine learning algorithms widely used in today's data-driven world. Among the models discussed are decision trees, Bayesian networks, principal component regression, support vector regression, neural networks, linear regression, and Gaussian regression. We will examine each model's unique features to determine which is best for our data.
This thesis analyzes data via visualization and correlation matrix to understand what type of model we need according to our data. However, the primary objective is to assess the practicality and restrictions of each of the preselected models.
A weighted decision matrix will assist us to make the comparative analysis and select the optimal model from the three preselected models (decision trees, Gaussian processes, and linear regressions). The findings of this study may be beneficial for scholars working in the areas of data science, statistics, and machine learning.