Sale Prediction Using Machine Learning Algorithms
Vu, Khuyen (2023)
2023
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2023061223577
https://urn.fi/URN:NBN:fi:amk-2023061223577
Tiivistelmä
This thesis is to investigate the sales prediction accuracy by trial and error from different machine learning models and study the accuracy of these models. We
perform predictions based on machine learning techniques on shops’ total sales with two main approaches: Least square regression, and Tree-based methods. We collect and analyze the Shopee data and use the two approaches. With the evaluation using two metrics Mean Absolute Errors (MAE) and Soft Interval
Accuracy (SIA), the model that gives the best results is random forests. We show that the regression shrinkage models can provide faster results with a trade-o↵ of slightly higher errors.
perform predictions based on machine learning techniques on shops’ total sales with two main approaches: Least square regression, and Tree-based methods. We collect and analyze the Shopee data and use the two approaches. With the evaluation using two metrics Mean Absolute Errors (MAE) and Soft Interval
Accuracy (SIA), the model that gives the best results is random forests. We show that the regression shrinkage models can provide faster results with a trade-o↵ of slightly higher errors.