Robotics 2
Machine Vision and AI: Features
Quiz
Your ID Code:
Question 3: Which of these is the second rule of selecting features?
Question 2: Which of these is the first rule of selecting features?
Question 1: True of false: using more features will always make your machine learning algorithm work better.
True
False
All features must be extracted from the same data stream.
All features must be representative of the same dimension.
The features must correctly identify all classes.
The feature must be possible to extract from the raw data.
Question 4: Which of these is the third rule of selecting features?
Select as few features as necessary to separate the data.
Question 5: What is meant by the term 'feature extraction'?
Select as many features as possible from the data available.
Select no more than 3 features so that the data may be viewed in a plot.
Select all features that can be extracted from the raw data.
Questions 7-10: The whole process of machine learning for machine vision has five steps, four of which are a part of the training process. Use the drop-down boxes to list these steps in the correct order.
The feature must be able to be extracted from an image.
The feature must be useful for distinguishing between classes.
The feature must allow the data to be clustered in few classes.
The feature must allow a neural network to learn the data.
As the number of features used decreases, the success of a machine learning algorithm decreases exponentially.
As the number of features used increases, the amount of training data needed increases exponentially.
Question 6: What is meant by the principle of the 'curse of dimensionality'?
Feature extraction is the process of finding values for each feature from the raw data.
Feature extraction is the process of using a machine learning algorithm to create a model of data.
Feature extraction is the process of determining how many features are needed to separate the data.
Feature extraction is the process of determining how many classes are represented within the raw data.
As the number of features used decreases, the success of a machine learning algorithm increases exponentially.
As the number of features used increases, the amount of training data needed decreases exponentially.
1
2
3
4
1
2
3
4
1
2
3
4
Image segmentation
1
2
3
4
Feature extraction
Machine learning
Present the algorithm with training images