An integrated framework combining first-principles calculations and machine learning was developed to predict gas-sensing performance. Key descriptors such as adsorption energy, adsorption distance, ...