The Development of a Matchmaking System Through The Use of Reinforcement Learning For Pet Match (PATCH)
Radisa Hussien Rachmadi and Maria Seraphina Astriani

Computer Science Department, School of Computing and Creative Arts, Bina Nusantara University, Jakarta, Indonesia 11480

Abstract

Pets are sometimes considered as an extended family- they deserve all the love and attention just as they show to their owners. During the Covid-19 pandemic, there was a surge in pet adoptions as research shows that pets can help with depression and loneliness. Now that the pandemic has slowly faded, and people are coming out of their homes more often, pets are often left neglected which is why it is needed to create a solution to cater to connecting pets for either breeding purposes or playdate purposes. The author and his team decided to develop a mobile application that will serve as a platform to connect pet owners through a matchmaking style application integrated with machine learning. Developed using Neural Networks and Proximal Policy Optimization (PPO) to predict user^s best possible matches based on the user^s swipes.

Keywords: reinforcement learning, pet, matchmaking, machine learning

Topic: Artificial Intelligence and Data Science