A regression model to predict house prices was fit but performs poorly on both training and test sets (high error on both). Which actions should the data scientist take to improve model accuracy? Select three.

The model is underfitting, indicated by poor performance on both training and test sets. Underfitting means the model is too simple to capture the underlying patterns in the data. To address underfitting: 1. Decrease the model's regularization strength: Regularization penalizes model complexity. Reducing its strength allows the model to become more complex and better fit the training data. 2. Increase the number of training examples used for fitting: While not a direct fix for underfitting, more data can help the model learn more robust patterns, especially if the current training set is too small or unrepresentative. 3. Increase the number of input features (add more relevant features): Providing the model with more relevant information can help it understand the relationships in the data better, leading to improved accuracy. Increasing the number of test examples only changes the evaluation set size, not the model's learning process. Decreasing the number of input features would further simplify the model, exacerbating underfitting.