Which solution will meet these requirements?
Perform one-hot encoding on every possible option for each question of the survey.
Perform binning on all the answers each respondent selected for each question.
Use Amazon Mechanical Turk to create categorical labels for each set of possible responses.
Use Amazon Textract to create numeric features for each set of possible responses.
Explanations:
One-hot encoding is suitable for categorical data, where each possible response option becomes a binary feature (0 or 1). This allows the logistic regression model to interpret multiple selections per question effectively, as each combination of responses is captured in the dataset.
Binning is typically used to group continuous data into discrete intervals. It does not effectively capture the multiple selections of categorical responses and can lead to loss of information regarding individual responses.
Using Amazon Mechanical Turk to create categorical labels is not a direct solution for representing responses in a dataset. It involves human labeling rather than encoding the multiple-choice answers into a format suitable for machine learning.
Amazon Textract is used for extracting text and data from scanned documents, not for encoding survey responses. It does not create numeric features for categorical survey data in the context of machine learning model training.
From my perspective, the answer is:
Perform one-hot encoding on every possible option for each question of the survey.