ALL

Question 1

What is MAE metric?

Answer 1

• Mean absolute error is an average absolute difference between the target and predicted values.
• It measures the average magnitude of the errors, the difference between a target and predicted value and a set of predictions.
• a smaller value indicates a higher-quality model.

Question 2

What is MAPE metric?

Answer 2

• Mean absolute percentage error is the average absolute percentage difference between the labels and the predicted values.
• This metric ranges between zero and infinity
• a lower value indicates a higher-quality model.
• MAPE is not shown if the target column contains any zero values, in this case MAPE is undefined.

Question 3

What is RMSE metric?

Answer 3

• root mean squared error
• RMSE is more sensitive to outliers than MAE
• a smaller value indicates a higher-quality model, zero represents a perfect predictor

Question 4

What is R squared metric?

Answer 4

• is the square of the Pearson correlation coefficient between the observed and predicted values
• ranges from zero to one, where a higher value indicates a higher-quality model

Question 5

What is RMSLE metric?

Answer 5

• root mean squared logarithmic error metric is similar to RMSE except that it uses the natural logarithm of the predicted and actual values plus one.
• metric ranges from zero to infinity, a lower value indicates a higher-quality model
• penalizes under-prediction more heavily than over-prediction.
• a good metric when you don’t want to penalize differences for large prediction values more heavily than for small prediction values.
• is returned only if all labeled and predicted values are non-negative.

Question 6

Why is it important to review how much each feature impacts a model?

Answer 6

Review this information to ensure that all the most important features make sense for your data and business problem.

Question 7

What is Log loss?

Answer 7

Log loss, this is the cross-entropy between the model predictions and the target values.

Question 8

When is Batch prediction used?

Answer 8

Batch prediction is useful for making many prediction requests at once.

Question 9

When are predicted values available in Batch prediction?

Answer 9

Batch prediction is asynchronous, meaning that the model will wait until it processes all of the prediction requests before returning a CSV file or a BigQuery Table with prediction values.

Question 10

When a model deployment is needed?

Answer 10

Deploy your model to make it available for online or real-time predictions.
- Predictions are available through requests using a REST API.
- Online prediction is synchronous or real-time, meaning the model only accepts one prediction request per API call.

Question 11

What are data requirements for Vertex AI AutoML?

Answer 11

• maximum data set size is 100 gigabytes
• the minimum row requirement is 1,000.
• the target column must be categorical or numerical.

Question 12

What is Google BigQuery ML?

Answer 12

Google BigQuery ML is a set of SQL extensions to support machine learning.

Question 13

What are the advantages of BigQuery ML?

Answer 13

• an easy to use way to invoke machine learning models on structured data using just SQL
• can provide decision-making guidance through predictive analytics
• speeds up the time to production,
• makes development work much easier,
• automates a number of the steps in the ML workflow.

Question 14

Which model types in BigQuery ML have hyperparameter tuning?

Answer 14

Hyperparameter tuning supports the following model types:
- linear regression,
- logistic regression,
- K means,
- matrix factorization,
- boosted tree classifier,
- boosted tree regressor,
- DNN classifier
- DNN regressor.

Question 15

What is TensorFlow Playground?

Answer 15

TensorFlow Playground is a powerful tool for visualizing how neural networks work.