Talk:Singular integral

Kernels or integral kernels

Despite User:Thenub314's edit summary, both terms are used. In fact, unless it is clear from context, the term "integral kernel" should be preferred because of possible confusion with the unrelated term kernel (mathematics). Springer's Encyclopedia of Mathematics uses the term "Kernel of an integral operator". I strongly feel that some initial context-setting is needed before the term kernel is simply tossed out there in the lead. Prefacing it with "integral" is one way to do this, albeit an imperfect one. silly rabbit (talk) 19:12, 30 April 2008 (UTC)[reply]

I have worked over the second and third sentence of the lead a bit to try to make it less awkward. It's still awkward, which probably means it needs to be rewritten, but I think it is an improvement nonetheless. silly rabbit (talk) 19:32, 30 April 2008 (UTC)[reply]

I didn't mean to offend. I do not recall having encountered the phrase "integral kernel." I agree it should be made very clear that it is a distinct concept from kernel (mathematics). This may even deserve a direct comment. It is not used, for example, in any of the reference works cited on the subject. I was just trying to keep the standard terminology. Kernel function is acceptable. Springer's choice of "Kernel of an integral operator" I feel is great. The lead could use a little work, just trying to help. Thenub314 (talk) 03:41, 1 May 2008 (UTC)[reply]

T(b) theorem.

There is something fishy about the statement of T(b) theorem. For one thing, we define coercive and use accretive. Also I seem to recall the paper we reference proves a more general theorem using para-accretive functions and requiring T(b) and T^t(b) to be functions in BMO. I will have to check this out a bit later. Thenub314 (talk) 11:19, 12 September 2008 (UTC)[reply]

Radius 10?

What is going on with this business of a normalised bump of radius 10? Where did the 10 come from? —Preceding unsigned comment added by 76.210.76.193 (talk) 05:18, 22 January 2009 (UTC)[reply]

There is nothing specific about 10. If I am not mistaken any absolute constant will do. Thenub314 (talk) 09:03, 22 January 2009 (UTC)[reply]

BMO

The link to BMO actually links to the Bank of Montreal. It needs some fixing!

Jose Brox

Thanks Jose! I have changed it to point to the correct place Thenub314 (talk) 12:45, 8 March 2009 (UTC)[reply]