Using the suffix "able", many verbs can be made into adjectives that describe what it is that the relevant (perhaps implicit) subject can perform or be allowed to do according to the verb. Examples:

  • understandable (from S can be understood)
  • destructible
  • learnable
  • incrementable

Now I am looking for a suffix that says that the subject is not just able to do this, but it must be done like that.

One way around it, would be to say the oposite and negate it. So for "must stay", we can use the anonym "destruct", make it the adjective "destructiable" and negate it:

  • indestructible, undestroyable, unbreakable, unwreckable

But I would prefer a syllable that adds the "must" at the end. So for example: There exist "incremental learning" and "batch learning" (learning on thing after the other, learning everything at once). These two words are jargon in machine learning.

So, now I want to say that some learning technique "enables learning one thing after the other", so I call it "incrementable learning" (the technique might also allow batch learning, but it also allows incremental learning).

Now I want to say a different technique enables to learn in batch, so is "batchable learning" (but it could also enable incrementale learning - this is not clarified!).

Now I want to say a third (different) technique only enables batch learning. No incremental learning is possible, so it is "batch-must learning"...? What is the right syllable - if there is one...

Looking for an antonym at https://en.wiktionary.org/wiki/incrementable did not help.


1 Answer 1


Short answer

Use -only as a suffix to the attributively used noun.

For instance:

  • incremental-only learning
  • batch-only learning


First of all, this is a somewhat strange question. In the area of computer science and artificial intelligence, there is a term called online machine learning. It includes several different types of algorithms for such learning, including incremental learning and batch learning.

The starting point of this question is that we can also use the words incrementable learning and batchable learning, to refer to the types of learning that are possible in certain situations.

But while incrementable is an understandably derived word using the normal rules of syntax, and Google Scholar shows 1,570 hits for it in scientific articles, it's not nearly as common as the simpler incremental. In comparison, Google Scholar shows 3,130,000 hits for incremental.

Comparing the two at Google Books Ngram Viewer, although incrementable exists, it is essentially a flat line at the bottom of a graph in comparison to incremental.

incrementable versus incremental

As uncommon as incrementable is, batchable is even less common.

Google Scholar shows only 329 hits for batchable in scientific articles as opposed to 4,150,000 hits for batch. Meanwhile, Google Books Ngram Viewer shows no results at all for batchable.

So, while both are understandable, and they have been used by some people, they are far from the common way of expressing such an idea.

It would be more common to rephrase the two:

  • incrementable learning → capable of incremental learning
  • batchable learning → capable of batch learning

Using -only

The use of only as an adjective is well defined:

2 a : alone in a class or category : SOLE
       // the only one left
       // the only known species
       // She's the only person you can really trust.

It can also be used as a modifier to an attribute noun that modifies another noun.

Depending on style and context, it can come before or after the main noun, and can be hyphenated, put in italics or quotes, or left just as it is:

  • I am a dogs-only veterinarian surgeon.
  • I am a "dogs only" veterinarian surgeon.
  • I am a dogs only veterinarian surgeon.
  • I am a veterinarian surgeon of dogs only.

But while using -only as a suffix for incremental and batch would technically work in the same way that incrementable and batchable would technically work (and I think it's the only reasonable way of satisfying what the question is looking for specifically), it suffers from the same problem as those other words: its use in this context is extremely uncommon.

I am unable to determine how to get Google Scholar to search only for hyphenated phrases (it ignores the hyphen), and I suspect it might not be possible. But, in a reversal of what I found with incrementable and batchable, Google Books Ngram Viewer shows no results at all for incremental-only, and it shows batch-only as a flat line at the bottom of a graph in comparison to batch.

batch versus batch-only

As such, as with incrementable and batchable, it would be more common to reprhase the expressions:

  • incremental-only learning → it can only learn incrementally
  • batch-only learning → it can only learn in a batch
  • Batch learning is not online learning. It is described in the wikipedia-article for online learning [1] to contrast it to online learning. "Incremental learning" in that article [1] is different to online learning, however in the incremental learning article [2] the term "incremental learning" is used as a synonym to "online learning". Afaik using those as synonyms is indeed how I discovered the term elsewhere. The terminology in machine learning is a mess.
    – Make42
    Jul 3, 2020 at 15:28
  • @Make42 I said online machine learning, and the article I linked to clearly shows both incremental learning and batch learning as being types of such machine learning. Just online learning (without the involvement of machines) is something else entirely and doesn't apply here. The terms incremental learning and batch learning have specific niche definitions in their use within machine learning that are not applicable in a more general sense. In fact, as mentioned in comments under the question, batch learning makes little sense outside of its machine learning context. Jul 3, 2020 at 15:36
  • Yes, "online learning" is just a common abbrevation of "online machine learning" in the field of machine learning. Of course I do not mean "online learning" as in "learn a different language in an online course". Sorry for the confusion that caused.
    – Make42
    Jul 3, 2020 at 15:46
  • I also linked to the same article with [1] above, so, I know which article you where linking to. There is a subsection "2.1 batch learning" in the article, but still, batch learning is no online machine learning (OML). Unfortunately, the article does not state that (it is not a great article). To realize it, you need to read the description of OML and the math for that section "2.1". Then you realize it is used to introduce the issues regarding online machine learning. A possibly more credible source: link.springer.com/chapter/10.1007/978-1-4842-4470-8_15
    – Make42
    Jul 3, 2020 at 15:53
  • All that being said, thanks for your post!
    – Make42
    Jul 3, 2020 at 15:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.