The regularisation approach can be used to remedy underfitting or overfitting concerns in statistical models. If some model parameters are likely to contribute to overfitting, regularisation techniques like LASSO can assist punish them.
You can suggest cross-validation techniques like k-folds cross-validation if the interviewer asks about alternative approaches to minimize overfitting.
Another option is to simplify the model by incorporating fewer variables and parameters. This reduces the amount of noise in the training data.