The use of LSTM-RNN model to forecast end-of-life clothing in Canadian thrift stores

An excessive amount of end-of-life (EOL) textile in Canadian thrift stores have been reported since the first wave of COVID-19 in March 2020. The donated clothes have gradually filled the thrift stores and a management plan is urgently needed. The quantity and composition of EOL textile, as well as the seasonal variations, are poorly understood. This study proposes a Recurrent Neural Network (RNN) model to examine EOL textile quantity and composition. Our partnering organization, MCC Canada, is one of the largest thrift chains in North America. Time series data in Saskatchewan stores will be used to train, validate, and test the RNN model. The final RNN model will then be used to forecast donated EOL textile temporally and spatially, improving the planning and operation of the local thrift industry. The proposed work will help the thrift industry to move closer to the establishment of a circular economy.

Faculty Supervisor:

Kelvin Tsun Wai Ng

Student:

Partner:

MCC Canada

Discipline:

Engineering

Sector:

Retail trade

University:

University of Regina

Program: