This repository contains the Public Thesis for my Master Degree in Computer Science. Due to NDA i'm not allowed to share the code.
Estimating shoe comfort from parametric foot shapes: a machine learning approach
This thesis aims to develop a machine learning model that, starting from a parametric foot shape and shoe shape, estimate the comfort of a particular shoe/size for a specific user. The comfort is catalogued in Tight Good and Loos in different foot local regions (such as Heel, Instep, Toe etc..) and a global Fit that represent the fit that the user perceive overall. The thesis travels across different methodologies both in model structure and data preprocessing with the aim of creating a good foundation for future researches and analysis.