Πανεπιστήμιο Μακεδονίας
Quality predictions with cross-sectional data from the Greek wine industry and clustering of industries with input-output data from the US economy using machine learning methods
Abstract
dc:descriptionThis thesis applies a range of statistical and machine learning models—implemented in R—to analyze data from e-commerce and the wine industry, focusing on predicting product quality and price, and on clustering economic sectors. It examines online markets and auctions using functional regression, models wine quality with linear and logistic regression as well as neural networks, and compares advanced methods such as LASSO, Random Forest, LDA, CART, kNN, and SVM, with Random Forest achieving up to 95% prediction accuracy. Finally, principal component analysis of U.S. input-output tables (2002–2019) identifies key industries and structural shifts, offering insights for economic policy and industrial dynamics.
Degree
thesis:*- Grantor dc:publisher
- Πανεπιστήμιο Μακεδονίας
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Athanasiadis, Ioannis
- Αθανασιάδης, Ιωάννης
Subjects
dc:subject × 20- Ποιοτικές προβλέψεις
- Επεξηγήσιμη μηχανική μάθηση
- Μη επιβλεπόμενη μάθηση
- Μη επιβλεπόμενη μέθοδος κύριων συνιστωσών
- Ανάλυση συστάδων
- Δίκτυα, Νευρωνικά τεχνητά
- Δενδρόγραμμα
- Quality predictions
- Supervised machine learning
- Unsupervised Machine Learning Algorithms
- Principal component analysis
- Clustering algorithms
- Αrtificial neural networks optimization
- Dendrograms
- Μαθηματικά
- Φυσικές Επιστήμες
- Στατιστική και Πιθανότητες
- Mathematics
- Natural Sciences
- Statistics and Probability
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/59991