Pet Adoption Status Prediction Based on Multiple Machine Learning Models

Honglin Lu

2024

Abstract

It is urgent and challenging to reasonably select machine learning algorithms and models to accurately predict the adoption status of pets. In this paper, the prediction is a classification task for whether a pet will be adopted or not which aims to reach the highest possible accuracy. Several models are used to predict the pet adoption status based on some information in the given dataset on Kaggle and the evaluation is mainly based on accuracy. For each model, pre-processing includes separating numerical and categorical columns and dropping useless columns or editing some columns if necessary. Multiple parameters are involved in tests for each model. The Artificial Neural Network (ANN) is designed to have four layers with the first three layers using ReLu as activation functions and the last using Sigmoid. Different amounts of middle layer neurons and epochs are tested in order to select the parameter with the highest accuracy to represent the model. By comparing the accuracy of each testing result, it indicates the best performing model is the four-layer ANN model with both the number of neurons in the middle layer and the amount of epoch to be around 60 and the number of neurons in the first and third layers to be 20.

Paper Citation

in Harvard Style

Lu H. (2024). Pet Adoption Status Prediction Based on Multiple Machine Learning Models. In Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM; ISBN 978-989-758-738-2, SciTePress, pages 362-367. DOI: 10.5220/0013331300004558

in Bibtex Style

@conference{mlscm24,
author={Honglin Lu},
title={Pet Adoption Status Prediction Based on Multiple Machine Learning Models},
booktitle={Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM},
year={2024},
pages={362-367},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013331300004558},
isbn={978-989-758-738-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM
TI - Pet Adoption Status Prediction Based on Multiple Machine Learning Models
SN - 978-989-758-738-2
AU - Lu H.
PY - 2024
SP - 362
EP - 367
DO - 10.5220/0013331300004558
PB - SciTePress