Which model is MOST likely to provide the best results in Amazon SageMaker?
Use the Amazon SageMaker k-Nearest-Neighbors (kNN) algorithm on the single time series consisting of the full year of data with a predictor_type of regressor.
Use Amazon SageMaker Random Cut Forest (RCF) on the single time series consisting of the full year of data.
Use the Amazon SageMaker Linear Learner algorithm on the single time series consisting of the full year of data with a predictor_type of regressor.
Use the Amazon SageMaker Linear Learner algorithm on the single time series consisting of the full year of data with a predictor_type of classifier.
Explanations:
The k-Nearest Neighbors (kNN) algorithm is not well-suited for time series forecasting, especially when limited to daily data from the past year. kNN is better for classification or regression tasks on static datasets, rather than predicting future values based on temporal trends.
The Random Cut Forest (RCF) algorithm is primarily designed for anomaly detection rather than direct forecasting. While it can analyze patterns in data, it does not provide a straightforward way to predict specific future values like air quality in parts per million.
The Linear Learner algorithm is appropriate for regression tasks and can effectively model the relationship in time series data, making it suitable for predicting future air quality based on historical daily data. It is capable of capturing trends and patterns over time.
The Linear Learner algorithm is suitable for regression, not classification in this context. Predicting air quality (a continuous variable) requires a regression approach rather than treating it as a classification problem, which would be inappropriate for forecasting ppm values.