Overfitting is a more frequent problem than underfitting and typically occurs as a result of trying to avoid overfitting.

Clear, visual explanation of the bias-variance tradeoff and how to find the sweet spot in your models. #BiasVariance #Overfitting #MachineLearningBasics ...

The fits shown exemplify underfitting (gray diagonal line, linear fit), reasonable fitting (black curve, third-order polynomial) and overfitting (dashed curve, fifth-order polynomial).

So, in summary, the key components of achieving good generalization (not underfitting or overfitting) in machine learning are hyper-parameter search, regularization, and out-of-sample testing.