@Sanniddha Chakrabarti Thanks, vector valued regression i.e. to treat the dependent variables as a vector.
Another approach is multi-task learning which means that you treat each dependent variable as a task but you learn the two tasks jointly together. In many cases this outperforms learning the tasks separately.
Please follow the Regression.