I am learning this book "Deep Learning and Convolutional Neural Networks for Medical Image Computing"

On the other hand, CNN models have been proved to have much higher modeling capacity, compared to the previous image recognition mainstream pipelines, e.g., HAAR, SIFT, HOG image features followed by spatial feature encoding, then random forest or support vector classifiers. Given millions of parameters to fit during model training (much more than previous pipelines), CNN representation empowers and enables computerized image recognition models, with a good possibility to be able to handle more challenging imaging problems. The primary risk is overfitting since model capacity is generally high in deep learning but often very limited datasets are available (that are with good quality of labels to facilitate supervised training). The core topics of this book are represented by examples on how to address this task-critical overfitting issue with deep learning model selection, dataset resampling and balancing, and the proper quantitative evaluation protocols or setups.

enter image description here

then indicates that the author put "CNN models" (method_1), "the previous image recognition mainstream pipelines" (method_2) and "spatial feature encoding" (method_3) in the comparison.

the question is

in the modeling capacity perspective, does then indicate method_1 > method_2 > method_3 ?

1 Answer 1


Yes, then does indicate order.

Then - (adverb) next or after that Cambridge Dictionary

In this context, it just means that the thing before the then comes before the thing after the then.

First, put on your socks, then put on your trousers, then put on your shoes.

He ate some salad, then some beef, and then finally he had a swig of beer.

  • Thanks for your answer. is the order method_1 > method_2 > method_3 what exactly what the author indicates by then?
    – user98358
    Commented Jul 15, 2019 at 9:42
  • 1
    I'm not familiar enough with the content to say for sure, but my interpretation would be: [HAAR, SIFT, HOG image features] -> [spatial feature encoding] -> [random forest or support vector classifiers].
    – Gamora
    Commented Jul 15, 2019 at 9:44
  • 1
    It's unusual for there to not be a conjunction before then. At least stylistically, I'd interpret the sentence as being HAAR, SIFT, HOG, or image features, followed by spatial feature encoding and then random forest or support vector classifiers. The use of and in this sentence would have avoided some of the confusion that's being caused by its absence. Commented Jul 15, 2019 at 13:45
  • @Jason +1 I would agree that it would be clearer if the author specified. I think they are probably assuming the reader knows a lot about the subject matter
    – Gamora
    Commented Jul 15, 2019 at 13:51

You must log in to answer this question.