Modeling Pet Compatibility: A Quantitative Approach to Evaluate Household Suitability for Pet Ownership

Throughout history, humans and animals have coexisted, relying on each other for aid and comfort. From offering companionship, to serving as hunting partners or providing other manual labor, our connection with animals runs deep. The study aims at developing mathematical models to assess whether households are suitable for having pets. The purpose is to provide a quantitative and objective approach for predicting pet adoption. For task 1, we first established an indicator system of three first-level indicators (Material Conditions, Environmental Conditions, and Mental Conditions) and 9 second-level indicators, then collected 20 sets of data from 20 organizations. The findings of this research affirm that the model can effectively distinguish between suitable and unsuitable households for keeping cats in Finland while it also applies elsewhere in diverse countries. It also successfully projects the number of households likely to adopt cats in the studied regions and is extended to assess the suitability for various pet types. In the United States, a remarkable 35100000 households have been projected as ready to adopt cats. Japan follows with an estimated 8400000 households prepared for cat ownership. In the Germany, there is a substantial potential with 10500000 households ready to adopt cats. For task 2, we adjust these weights to accommodate four additional different pets (e.g., dogs, small mammals, ornamental fish, and birds), then the weights should be reevaluated in light of each pet’s unique needs and suitability for different home environments. The prediction using the random forest method yielded an F1 value of 81.59%. In addition, we perform quantitative estimation of the optimization model for multiple pets, and the obtained model works well. For task 3, linear regression and gray prediction fusion model are employed for predicting future pet ownership patterns. Using this forecast model, we predict pet ownership and retention trends over the next 5, 10, and 15 years. Despite this decrease, the retention levels for ornamental fish and parrots show relative stability in comparison to the more pronounced drops seen in the retention of hamsters, dogs, and cats. Finally, we conduct sensitivity analysis on our models. In the evaluation model, we changed the fluctuation percentage of the modulus, and the deviation fluctuation was within. Therefore, the model is robust to changes in a single indicator. In the predictive model, we changed the degree of intervention factors and obtained a small fluctuation deviation, proving that our predictive model also has good robustness.