How to correctly understand the recognition results on the MNIST website?

Question

Neural networks

The MNIST website published the results of digit pattern recognition by various algorithms. I have a question about the column: TEST ERROR RATE (%) . In this column, the higher the value, then the more errors were in the recognition? So the best classification model is the one with the lowest TEST ERROR RATE ?
Screenshot:

Link to MNIST classification results:
The MNIST database of handwritten digits

Answer 1

TEST ERROR RATE = erroneous recognition on the test set. So yes, the higher this value, the worse the given algorithm works.