Size fitting is a significant problem for online garment shops. The return rates due to size misfit are very high. We propose an ensemble (with an original and novel definition of the weights) of ordered logistic regression and random forest (RF) for solving the size matching problem, where ordinal data should be classified. These two classifiers are good candidates for combined use due to their complementary characteristics. A multivariate response (an ordered factor and a numeric value assessing the fit) was considered with a conditional random forest. A fit assessment study was carried out with 113 children. They were measured using a 3D body scanner to obtain their anthropometric measurements. Children tested different garments of different sizes, and their fit was assessed by an expert. Promising results have been achieved with our methodology. Two new measures have been introduced based on RF with multivariate responses to gain a better understanding of the data. One of them is an intervention in prediction measure defined locally and globally. It is shown that it is a good alternative to variable importance measures and it can be used for new observations and with multivariate responses. The other proposed tool informs us about the typicality of a case and allows us to determine archetypical observations in each class.Keywords: Multivariate conditional random forest; Proportional odds logistic regression; Supervised learning; Ordinal classification; Childrenswear garment fitting; Variable importance ACKNOWLEDGEMENTS This work has been partially supported by Grants DPI2013-47279-C2-1-R and DPI2013-47279-C2-2-R.An ensemble of ordered logistic regression and random forest for child garment size matchingSize fitting is a significant problem for online garment shops. The return rates due to size misfit are very high. We propose an ensemble (with an original and novel definition of the weights) of ordered logistic regression and random forest (RF) for solving the size matching problem, where ordinal data should be classified. These two classifiers are good candidates for combined use due to their complementary characteristics. A multivariate response (an ordered factor and a numeric value assessing the fit) was considered with a conditional random forest. A fit assessment study was carried out with 113 children. They were measured using a 3D body scanner to obtain their anthropometric measurements. Children tested different garments of different sizes, and their fit was assessed by an expert. Promising results have been achieved with our methodology. Two new measures have been introduced based on RF with multivariate responses to gain a better understanding of the data. One of them is an intervention in prediction measure defined locally and globally. It is shown that it is a good alternative to variable importance measures and it can be used for new observations and with multivariate responses. The other proposed tool informs us about the typicality of a case and allows us to determine archetypical obs...
Today, there is an increasing availability of human body 3D data and an increasing number of anthropometric owners. This is due to the fact of the progressive conduction of large national surveys using high resolution 3D scanners and due to the increasing number of low-cost technologies for acquiring body shape with electronic consumer devices like webcams, smartphones or Kinect. However, the commercial use and exploitation in industry of digital anthropometric data is still limited to the use of 1D measurements extracted from this vast 3D information. There is a lack of universal resources enabling: to conjointly use and analyse datasets regardless from the source or type of scanning technology used, the flexible measurement extraction beyond pre-defined sets, and the analysis of the information contained in human shapes. This paper presents four software tool solutions aimed at addressing different user profiles and needs regarding the use and exploitation of the increasing number of 3D anthropometric data
This paper presents partial results of a larger validation study of different Data-driven 3D Reconstruction (D3DR) technologies developed by IBV to create watertight 3D human models from measurements (1D3D), 2D images (2D3D) or raw scans (3D3D). This study quantifies the reliability (Standard Error of Measurement, SEM; Mean Absolute Deviation, MAD; Intra-class Correlation Coefficient, ICC; and Coefficient of Variation, CV) of body measurements taken on human subjects. Our results are also compared to similar studies found in literature assessing the reliability of digital and traditional anthropometry. Moreover, we assess the compatibility (bias and Mean Absolute Error, MAE) of measurements between D3DR technologies. The results show that 2D3D can provide visually accurate body shapes and, for the measurements assessed, 2D3D is as reliable as high resolution 3D scanners. It is also more accurate than manual measurements taken by untrained users. Due to accessibility, cost and portability (e.g. 2D3D built in a smartphone app) they could be more suitable than other methods at locations where body scanners are not available such as homes, medical or physical therapy offices, and small retail stores and gyms.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.