367:. The unscaled sum of iid variables doesn't converge to anything at all! Of course, if you take a distribution with support on the non-negative integers, and multiply it by a non-integer scalar, the resulting distribution will not be supported by the non-negative integers. But there is an alternate re-scaling operation which does yield an integer-supported distribution, and that has a central limit theorem that converges to a Poisson distribution (or more generally, a discrete-stable distribution, if the finite variance assumption is dropped). --
170:
152:
116:
22:
64:
78:
53:
342:
Poisson distributed, and approaches a normal distribution in the limit, in a certain sense. Perhaps we should mention the discrete limit theorems are special cases of the continuous ones? Some care is needed to indicate which limits we are taking, and what kind of convergence. Random sums of
Poisson
343:
RVs can themselves approach non-Poisson Stable RVs in the limit, according to the linked reference from Lee, for example. Maybe here's not the case to explain that? In whcih case we should remove the section entirely again; currently it's ambiguous what sort of limits we are discussing etc. --
275:
That is not the case. Under the
Central Limit Theorem, the normal distribution is also the limit a discrete distribution with finite variance. The CLT is not dependent on discrete vs. continuous. A binomial and even Poisson itself tends to the normal.
486:
The discrete-stable distributions are a class of probability distributions with the property that the sum of several random variables from such a distribution is distributed according to the same family....
491:
The most well-known discrete stable distribution is the
Poisson distribution which is a special case as the only discrete-stable distribution for which the mean and all higher-order moments are finite.
238:
Something must be missing from the description of limit theorems, since the sum of many discrete iid random variables converges to a normal distribution, which is not a discrete-stable distribution.
358:
So now I understand better what the article is getting it. The deleted section was not wrong, but there was very important context missing from the article, hopefully I will get around to fixing it.
216:
89:. The project works to allow users to contribute quality articles and media files to the encyclopedia and track their progress as they are developed. To participate, please visit the
499:
Am I misunderstanding "stable" in this context? Or is there supposed to be a difference between "discrete stable" and "discrete-stable"? Or am I missing something else? Help? --
338:
Some of this confusion is perhaps due to the fact that in the limit one can approach, in distribution, both continuous and discrete RVs; Indeed a sum of independent
Poisson RVs
316:
I removed the section on limit theorems. I am not finding any support for discrete stable distributions satisfying this property except for lattice random variables. In fact
460:
418:
558:
210:
543:
90:
85:
58:
323:(stable distribution with index of stability Ī± = Ī³ and with skewness parameter Ī² = 1), whose Laplace transform is given by exp(āĪ»tĪ³ )." (emphasis added)
291:
563:
553:
186:
134:
548:
538:
177:
157:
319:
explicitly states that at least some discrete stable distributions belong "to the domain of normal attraction of the absolutely
519:
504:
372:
496:
The difference of two independent
Poisson variables is discrete, stable, and has all moments finite, but it is not Poisson.
363:
Briefly: it is usually left unsaid, but actually, when we invoke the central limit theorem what we really mean is a sum
33:
256:
and has finite variance, the limit is the normal distribution as you point out. In the case that the distribution is
515:
500:
368:
128:
63:
470:
348:
21:
295:
39:
169:
151:
124:
466:
344:
185:
on
Knowledge. If you would like to participate, please visit the project page, where you can join
423:
328:
303:
281:
265:
243:
317:
393:
523:
508:
474:
376:
352:
332:
307:
285:
269:
247:
532:
324:
299:
277:
261:
239:
252:
I have tried to make it more explicit that in the case when the distribution is
182:
115:
514:
As I indicate above, there is important context missing from the article. --
77:
52:
15:
114:
260:
has finite mean, the limit is the
Poisson distribution.
123:
This article was accepted on 19 March 2015 by reviewer
426:
396:
386:
Generating probability
Generating Function definition
181:, a collaborative effort to improve the coverage of
454:
412:
215:This article has not yet received a rating on the
462:. We should either switch notation, or clarify.
8:
99:Knowledge:WikiProject Articles for creation
19:
146:
102:Template:WikiProject Articles for creation
83:This article was reviewed by member(s) of
47:
446:
425:
404:
395:
390:We link to a PGF definition in terms of
321:continuous positive stable distribution
148:
49:
559:Unknown-importance Statistics articles
420:for a RV X, but then work in terms of
544:AfC submissions by date/19 March 2015
7:
175:This article is within the scope of
38:It is of interest to the following
14:
86:WikiProject Articles for creation
195:Knowledge:WikiProject Statistics
168:
150:
76:
62:
51:
20:
564:WikiProject Statistics articles
554:Start-Class Statistics articles
198:Template:WikiProject Statistics
443:
430:
290:We have illustrations of this
1:
524:21:58, 23 December 2016 (UTC)
509:18:54, 14 December 2016 (UTC)
377:21:58, 23 December 2016 (UTC)
189:and see a list of open tasks.
475:08:00, 6 December 2015 (UTC)
353:08:11, 6 December 2015 (UTC)
580:
455:{\displaystyle E(1-s)^{X}}
333:03:09, 17 April 2015 (UTC)
308:02:34, 17 April 2015 (UTC)
286:02:31, 17 April 2015 (UTC)
270:20:01, 16 April 2015 (UTC)
248:15:26, 28 March 2015 (UTC)
214:
163:
122:
71:
46:
549:Accepted AfC submissions
539:Start-Class AfC articles
456:
414:
413:{\displaystyle Es^{X}}
178:WikiProject Statistics
119:
28:This article is rated
516:God made the integers
501:God made the integers
457:
415:
369:God made the integers
118:
96:Articles for creation
93:for more information.
59:Articles for creation
424:
394:
201:Statistics articles
452:
410:
120:
34:content assessment
231:
230:
227:
226:
223:
222:
145:
144:
141:
140:
571:
461:
459:
458:
453:
451:
450:
419:
417:
416:
411:
409:
408:
217:importance scale
203:
202:
199:
196:
193:
172:
165:
164:
154:
147:
107:
106:
103:
100:
97:
80:
73:
72:
67:
66:
65:
55:
48:
31:
25:
24:
16:
579:
578:
574:
573:
572:
570:
569:
568:
529:
528:
482:
442:
422:
421:
400:
392:
391:
388:
236:
200:
197:
194:
191:
190:
125:GraemeĀ Bartlett
104:
101:
98:
95:
94:
61:
32:on Knowledge's
29:
12:
11:
5:
577:
575:
567:
566:
561:
556:
551:
546:
541:
531:
530:
527:
526:
494:
493:
488:
481:
478:
467:Livingthingdan
449:
445:
441:
438:
435:
432:
429:
407:
403:
399:
387:
384:
382:
380:
379:
360:
359:
345:Livingthingdan
336:
335:
313:
312:
311:
310:
235:
234:Limit theorems
232:
229:
228:
225:
224:
221:
220:
213:
207:
206:
204:
187:the discussion
173:
161:
160:
155:
143:
142:
139:
138:
121:
111:
110:
108:
81:
69:
68:
56:
44:
43:
37:
26:
13:
10:
9:
6:
4:
3:
2:
576:
565:
562:
560:
557:
555:
552:
550:
547:
545:
542:
540:
537:
536:
534:
525:
521:
517:
513:
512:
511:
510:
506:
502:
497:
492:
489:
487:
484:
483:
479:
477:
476:
472:
468:
463:
447:
439:
436:
433:
427:
405:
401:
397:
385:
383:
378:
374:
370:
366:
362:
361:
357:
356:
355:
354:
350:
346:
341:
334:
330:
326:
322:
318:
315:
314:
309:
305:
301:
297:
293:
289:
288:
287:
283:
279:
274:
273:
272:
271:
267:
263:
259:
255:
250:
249:
245:
241:
233:
218:
212:
209:
208:
205:
188:
184:
180:
179:
174:
171:
167:
166:
162:
159:
156:
153:
149:
136:
133:
130:
126:
117:
113:
112:
109:
92:
88:
87:
82:
79:
75:
74:
70:
60:
57:
54:
50:
45:
41:
35:
27:
23:
18:
17:
498:
495:
490:
485:
464:
389:
381:
365:and re-scale
364:
339:
337:
320:
257:
253:
251:
237:
176:
131:
105:AfC articles
91:project page
84:
40:WikiProjects
30:Start-class
533:Categories
254:continuous
192:Statistics
183:statistics
158:Statistics
258:discrete
135:contribs
325:Rlendog
300:Rlendog
278:Rlendog
262:Wainson
240:Rlendog
36:scale.
480:What?
520:talk
505:talk
471:talk
373:talk
349:talk
329:talk
304:talk
296:here
294:and
292:here
282:talk
266:talk
244:talk
129:talk
211:???
535::
522:)
507:)
473:)
465:--
437:ā
375:)
351:)
340:is
331:)
306:)
298:.
284:)
268:)
246:)
137:).
518:(
503:(
469:(
448:X
444:)
440:s
434:1
431:(
428:E
406:X
402:s
398:E
371:(
347:(
327:(
302:(
280:(
264:(
242:(
219:.
132:Ā·
127:(
42::
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.