Which of the following is used to overcome from the underfitting?Use data augmentation techniqueRemove outliers in the training setAdd more features to the dataSelect a model with lesser features

To overcome underfitting, there are several techniques that can be used. Let's go through each option and see if it helps in addressing underfitting:

1. Use data augmentation technique: Data augmentation involves creating new training examples by applying various transformations to the existing data. This can help in increasing the size and diversity of the training set, which in turn can improve the model's ability to generalize. Data augmentation can be particularly useful when the training set is small or lacks diversity.

2. Remove outliers in the training set: Outliers are data points that deviate significantly from the rest of the data. They can have a negative impact on the model's performance, especially if the model is sensitive to extreme values. By removing outliers from the training set, we can reduce their influence on the model and potentially improve its ability to fit the data.

3. Add more features to the data: Underfitting can occur when the model is too simple and lacks the necessary complexity to capture the underlying patterns in the data. By adding more features to the data, we can provide the model with more information to learn from, potentially improving its ability to fit the data more accurately.

4. Select a model with lesser features: This option seems counterintuitive in the context of overcoming underfitting. Typically, underfitting occurs when the model is too simple and lacks the necessary complexity to capture the underlying patterns in the data. Selecting a model with lesser features may exacerbate the underfitting problem. Therefore, this option is not recommended for overcoming underfitting.

In summary, to overcome underfitting, it is generally recommended to use data augmentation techniques, remove outliers from the training set, and/or add more features to the data.