Which approach should a Machine Learning Specialist take to obtain accurate predictions?
Prepare the images in RecordIO format and upload them to Amazon S3. Use Amazon SageMaker to train, test, and validate the model using an image classification algorithm to categorize images into various weed classes.
Prepare the images in Apache Parquet format and upload them to Amazon S3. Use Amazon SageMaker to train, test, and validate the model using an object- detection single-shot multibox detector (SSD) algorithm.
Prepare the images in RecordIO format and upload them to Amazon S3. Use Amazon SageMaker to train, test, and validate the model using an object- detection single-shot multibox detector (SSD) algorithm.
Prepare the images in Apache Parquet format and upload them to Amazon S3. Use Amazon SageMaker to train, test, and validate the model using an image classification algorithm to categorize images into various weed classes.
Explanations:
This option suggests using an image classification algorithm, which categorizes entire images rather than detecting objects within them. Since the requirement is to detect specific types of weeds and their locations, an object detection approach is necessary.
While this option correctly suggests using an object detection algorithm, it incorrectly specifies Apache Parquet format for the images. RecordIO is the appropriate format for image datasets in Amazon SageMaker.
This option proposes using the appropriate RecordIO format and an object detection algorithm (SSD). This aligns with the requirement to detect and locate specific types of weeds in the images captured by the tractor-mounted cameras.
Similar to option A, this option suggests using an image classification algorithm, which does not meet the need for detecting the location of specific weeds. Additionally, Apache Parquet is not the suitable format for this task.
I draft that the answer is:
Prepare the images in RecordIO format and upload them to Amazon S3. Use Amazon SageMaker to train, test, and validate the model using an object- detection single-shot multibox detector (SSD) algorithm.