Question-61: Why overfitting makes model useless with the new data?

Answer: Overfitting happens when your model performs better with the training data and have more or high accuracy with the training data set. However, its accuracy drops when new data is given to the model. Because model is not generalized for new data and overfit with the training data set.

Question-62: How can we avoid overfitting in Machine Learning Model?

Answer: To prevent overfitting with the Machine Learning Model we can do following things

• Increase the volume of data for training: If you have more data to train your model then its accuracy increases.

There are more ways, we will discuss in later Questions as you learn more concepts.

Question-63: What do you mean by Cross Validation in Machine Learning?

Answer: We use cross validation to detect model overfitting or finding whether model is failed to generalize itself. Using the cross-validation technique, we can evaluate the various Machine Learning Models using the sample test data. And later validate with other sample data to find the accuracy of the model.

Question-64: What is the model underfitting?

Answer: Your model is called underfit when it performs badly with the training data itself.

Question-65: Why underfitting happens on the Machine Learning Model?

Answer: Underfitting happens because the model you are working upon is not able to capture the relationship between the input examples and the target variables.

Y = f(X)

X=Input Variable.

Y= Output variable.