Alpine skis categorization with comprehensive decision tree models

Skiers are faced with an ever-growing number of alpine skis to choose from. To simplify the shopping process, skis are divided into categories based on their intended usage. However, these categories are loosely defined, with numerous categorizations available. This paper presents a methodology base...

Full description

Saved in:
Bibliographic Details
Published in:Sports engineering 2023-12, Vol.26 (1), Article 12
Main Authors: Benghanem, Abdelghani, Audet, Jonathan, Lussier-Desbiens, Alexis
Format: Article
Language:English
Subjects:
Biomedical Engineering and Bioengineering
Categories
Classification
Decision trees
Engineering
Engineering Design
Materials Science
Original Article
Rehabilitation Medicine
Retail sales
Retail stores
Shopping
Skis
Sports Medicine
Theoretical and Applied Mechanics
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Skiers are faced with an ever-growing number of alpine skis to choose from. To simplify the shopping process, skis are divided into categories based on their intended usage. However, these categories are loosely defined, with numerous categorizations available. This paper presents a methodology based on decision trees to investigate how alpine skis are currently classified by the ski industry. As an example, simple, robust, and clear classification rules have been identified from the categories used by a large retailer. The rules found are only based on the measured physical attributes of the skis. Decision trees identified a few classification rules, three or fewer in all but one case, that provide high accuracy (88–97.5%) and inclusiveness (85–100%, except for the All-Mountain category). The rules found are easy to interpret by most skiers, especially when analyzing the other attributes highly correlated with each splitting rule of the decision trees. The rules found could be used by retailers to classify skis from different brands uniformly and automatically. Such rules could also be used to educate customers and simplify their shopping experience.
ISSN:1369-7072
1460-2687
DOI:10.1007/s12283-023-00405-9