You work as a machine learning specialist for a consulting firm that has the NFL as a client.
You are working on the passer completion probability model using statistics from in-play metrics.
You are running your linear learner model in Amazon SageMaker using a CSV file representation of your passer completion probability statistics.
You are now running your inference. Some of the features and their data types are listed below. |Feature Name |Data Type | | Passer age |Numeric| | Length of pass|Numeric| | Complete (yes/no)|Categorical | |Feature Name|Data Type | | Distance between receiver and nearest defender|Numeric| | Play called (post, crossing, screen, etc.)|Categorical | You are using the Complete feature as your prediction response feature.
You are now making predictions on new data.
When you interrogate the response of your model, which of the following do you expect to find?
Click on the arrows to vote for the correct answer
A. B. C. D.Answer: C.
Option A is incorrect.
For a binary classification (complete yes or no), the model produces a score denoting the strength of the prediction AND a predicted_label denoting complete or not complete.
Option B is incorrect.
This option describes the response for multiclass classification, but you are working with binary classification.
Option C is correct.
For a binary classification (complete yes or no), the model produces a score denoting the strength of the prediction AND a predicted_label denoting complete or not complete.
Option D is incorrect.
For a binary classification (complete yes or no), the model produces a score denoting the strength of the prediction AND a predicted_label denoting complete or not complete.
Reference:
Please see the Amazon SageMaker developer guide titled Linear Learner Algorithm.
As a machine learning specialist, you have built a linear learner model in Amazon SageMaker to predict the completion probability of a pass in the NFL game based on certain features like passer age, length of pass, distance between receiver and nearest defender, and play called. The prediction response feature that you have used is the "Complete" feature which is categorical and can have two possible values - "yes" or "no".
When you are making predictions on new data, the response of your model will contain information about the predicted outcome for each observation. In the case of a binary classification problem like this, the model will predict the probability of the positive class (i.e., "yes" for completion) and the negative class (i.e., "no" for non-completion).
So, when you interrogate the response of your model, you can expect to find the following information:
A. Score: The prediction produced by the model - This is the probability of the positive class (i.e., "yes" for completion). It is a single floating-point number between 0 and 1.
B. Score: The prediction produced by the model AND predicted_class which is an integer from 0 to num_classes-1 - This is the same as option A but also includes the predicted class label. In a binary classification problem like this, there are two possible classes (i.e., "yes" or "no"). Therefore, the predicted_class will be either 0 or 1.
C. Score: A single floating-point number measuring the strength of the prediction AND predicted_label which is 0 or 1 - This option is similar to option B but instead of an integer predicted_class label, it includes a binary predicted_label. The predicted_label will be 0 if the predicted probability of the positive class is less than 0.5 and 1 if it is greater than or equal to 0.5.
D. Score: The prediction produced by the model OR predicted_label which is 0 or 1 - This option is incorrect as it suggests that either the prediction score or the predicted label will be returned. In reality, the model will provide both the prediction score (i.e., the probability of the positive class) and the predicted label (either 0 or 1).
Therefore, the correct answer to the question is either A, B or C depending on how much detail is required in the response. Option A provides only the prediction score, Option B provides the prediction score and the predicted class label, and Option C provides the prediction score and a binary predicted label.