Knowledge (XXG)

ZPP (complexity)

Source 📝

90: 25: 621:
By repeated random selection of a possible witness, the large probability that a random string is a witness gives an expected polynomial time algorithm for accepting or rejecting an input. Conversely, if the Turing Machine is expected polynomial-time (for any given x), then a considerable fraction of
577:
with a second input tape for V on which is written the sequence of coin flips. By selecting the witness as a random string, the verifier is a probabilistic polynomial-time Turing Machine whose probability of accepting x when x is in
353:, the chance that it will yield an answer before we stop it is at least 1/2. This means the chance we'll give the wrong answer on a YES instance, by stopping and yielding NO, is at most 1/2, fitting the definition of an 481:
The witness need not be a traditionally construed proof. If V is a probabilistic Turing machine which could possibly accept x if x is in X, then the proof is the string of coin flips which leads the machine to accept
289:
Given an input, run A on the input for one step. If it returns YES, the answer must be YES. Otherwise, run B on the input for one step. If it returns NO, the answer must be NO. If neither occurs, repeat this
294:
Note that only one machine can ever give a wrong answer, and the chance of that machine giving the wrong answer during each repetition is at most 50%. This means that the chance of reaching the
675:
for itself, meaning that a ZPP machine with the power to solve ZPP problems instantly (a ZPP oracle machine) is not any more powerful than the machine without this extra power. In symbols,
475:
The definition is very asymmetric. The proof of x being in X is a single string. The proof of x not being in X is the collection of all strings, none of which is a proof of membership.
454:
can be thought of as the proof of membership. In the case of short proofs (of length bounded by a polynomial in the size of the input) which can be efficiently verified (
54: 152:
In other words, if the algorithm is allowed to flip a truly-random coin while it is running, it will always return the correct answer and, for a problem of size
862: 1259: 656:
language has V(x,w) accept on a (clear) majority of strings w if x is in X, and conversely reject on a (clear) majority of strings w if x is not in
213: 513:
a polynomial, only one need cause the verifier to accept (if x is in X. If x is not in X, no string will cause the verifier to accept).
1224: 76: 855: 535:
is the class of sets for which any random string is likely to be a witness of x in X, or x not in X, which ever the case may be.
122: 345:
its expected running time. If it gives an answer, give that answer. If it doesn't give any answer before we stop it, give NO.
531:
is the class of sets for which, if x is not in X, any randomly chosen string is likely to be a witness for non-membership.
184: 138: 37: 848: 211:
is based on probabilistic Turing machines, but, for clarity, note that other complexity classes based on them include
1047: 47: 41: 33: 1240: 660:. No single string w need be definitive, and therefore they cannot in general be considered proofs or witnesses. 58: 1229: 1183: 796: 350: 1178: 1173: 200:
It returns DO NOT KNOW with probability at most 1/2 for every input (and the correct answer otherwise).
622:
the runs must be polynomial-time bounded, and the coin sequence used in such a run will be a witness.
1168: 271: 173: 89: 333:, suppose we have a Las Vegas algorithm C to solve a problem. We can then construct the following 1188: 1152: 1042: 974: 871: 809: 672: 230: 134: 98: 1077: 509:
requires only that witnesses exist. They may be very rare. Of the 2 possible strings, with
1147: 1137: 1094: 1084: 1067: 1057: 1015: 1010: 1005: 989: 969: 947: 942: 937: 925: 920: 915: 910: 816: 247: 219: 130: 106: 94: 486:(provided all members in X have some witness and the machine never accepts a non-member). 1142: 1030: 952: 905: 830: 740: 303: 110: 1253: 172:), even though it might occasionally be much longer. Such an algorithm is called a 756:, i.e., every Las Vegas algorithm has a deterministic polynomial-time equivalent. 835: 489:
The co-concept is a proof of non-membership, or membership in the complement set.
1020: 930: 1132: 957: 527:
The corresponding co-classes have witness for non-membership. In particular,
1127: 1122: 1117: 1062: 885: 668:
It is known that ZPP is closed under complement; that is, ZPP = co-ZPP.
1112: 775: 765: 1072: 1219: 1214: 1089: 840: 636:
does not require witnesses, although witnesses are sufficient (hence
114: 361:
algorithm is identical, except that it gives YES if C "times out".
1209: 1204: 1025: 561:) corresponds to the deterministic polynomial-time Turing Machine 88: 1037: 895: 614:
is large, but zero of incorrectly accepting or rejecting x (for
844: 458:
is a polynomial-time deterministic Turing machine), the string
1052: 984: 979: 900: 890: 478:
For all x in X there must be a witness to its membership in X.
225: 197:
The answer is always either DO NOT KNOW or the correct answer.
148:
The running time is polynomial in expectation for every input.
102: 18: 381:
can be thought of in terms of proof of membership in a set.
93:
ZPP in relation to other probabilistic complexity classes (
594:); of rejecting x when x is not in X is large but zero if 550:
is easy. The probabilistic polynomial-time Turing Machine
259:. To show this, first note that every problem which is in 505:
are sets which have witnesses for membership. The class
117:. It is unknown if any of these containments are strict. 164:) such that the average running time will be less than 524:
any string chosen at random will likely be a witness.
748:, and some computer scientists have conjectured that 538:
Connecting this definition with other definitions of
282:
algorithm A and the (possibly completely different)
278:
Suppose we have a language L recognized by both the
245:
is exactly equal to the intersection of the classes
183:
can be defined as the class of problems for which a
1197: 1161: 1105: 998: 878: 229:is based on another machine with randomness: the 46:but its sources remain unclear because it lacks 145:It always returns the correct YES or NO answer. 582:is large (greater than 1/2, say), but zero if 255:. This is often taken to be the definition of 856: 391:V for a set X is a Turing machine such that: 8: 306:running time is polynomial. This shows that 194:It returns an answer YES, NO or DO NOT KNOW. 606:); and of correctly accepting or rejecting 863: 849: 841: 759:There exists an oracle relative to which 77:Learn how and when to remove this message 7: 204:The two definitions are equivalent. 573:) by replacing the random tape of 298:th round shrinks exponentially in 191:It always runs in polynomial time. 14: 1225:Probabilistically checkable proof 1260:Probabilistic complexity classes 23: 727:is obviously contained in both 664:Complexity-theoretic properties 187:exists with these properties: 141:exists with these properties: 1: 185:probabilistic Turing machine 139:probabilistic Turing machine 707:Connection to other classes 403:then there exists a string 156:, there is some polynomial 16:Concept in computer science 1276: 1241:List of complexity classes 628:should be contrasted with 129:(zero-error probabilistic 1238: 1230:Interactive proof system 137:of problems for which a 32:This article includes a 430:, then for all strings 237:Intersection definition 61:more precise citations. 1184:Arithmetical hierarchy 797:time hierarchy theorem 118: 1179:Grzegorczyk hierarchy 1174:Exponential hierarchy 1106:Considered infeasible 92: 1169:Polynomial hierarchy 999:Suspected infeasible 109:), which generalise 1198:Families of classes 879:Considered feasible 351:Markov's Inequality 341:Run C for at least 302:, showing that the 272:Las Vegas algorithm 174:Las Vegas algorithm 872:Complexity classes 207:The definition of 119: 34:list of references 1247: 1246: 1189:Boolean hierarchy 1162:Class hierarchies 779:would imply that 365:Witness and proof 123:complexity theory 87: 86: 79: 1267: 865: 858: 851: 842: 744:is contained in 699:is contained in 516:For the classes 325:is contained in 314:is contained in 231:quantum computer 135:complexity class 82: 75: 71: 68: 62: 57:this article by 48:inline citations 27: 26: 19: 1275: 1274: 1270: 1269: 1268: 1266: 1265: 1264: 1250: 1249: 1248: 1243: 1234: 1193: 1157: 1101: 1095:PSPACE-complete 994: 874: 869: 826: 805: 709: 666: 610:as a member of 555: 367: 357:algorithm. The 239: 179:Alternatively, 131:polynomial time 83: 72: 66: 63: 52: 38:related reading 28: 24: 17: 12: 11: 5: 1273: 1271: 1263: 1262: 1252: 1251: 1245: 1244: 1239: 1236: 1235: 1233: 1232: 1227: 1222: 1217: 1212: 1207: 1201: 1199: 1195: 1194: 1192: 1191: 1186: 1181: 1176: 1171: 1165: 1163: 1159: 1158: 1156: 1155: 1150: 1145: 1140: 1135: 1130: 1125: 1120: 1115: 1109: 1107: 1103: 1102: 1100: 1099: 1098: 1097: 1087: 1082: 1081: 1080: 1070: 1065: 1060: 1055: 1050: 1045: 1040: 1035: 1034: 1033: 1031:co-NP-complete 1028: 1023: 1018: 1008: 1002: 1000: 996: 995: 993: 992: 987: 982: 977: 972: 967: 962: 961: 960: 950: 945: 940: 935: 934: 933: 923: 918: 913: 908: 903: 898: 893: 888: 882: 880: 876: 875: 870: 868: 867: 860: 853: 845: 839: 838: 831:Complexity Zoo 825: 824:External links 822: 821: 820: 813: 804: 801: 769:. A proof for 708: 705: 665: 662: 553: 491: 490: 487: 479: 476: 448: 447: 420: 366: 363: 347: 346: 292: 291: 287: 238: 235: 202: 201: 198: 195: 192: 150: 149: 146: 85: 84: 42:external links 31: 29: 22: 15: 13: 10: 9: 6: 4: 3: 2: 1272: 1261: 1258: 1257: 1255: 1242: 1237: 1231: 1228: 1226: 1223: 1221: 1218: 1216: 1213: 1211: 1208: 1206: 1203: 1202: 1200: 1196: 1190: 1187: 1185: 1182: 1180: 1177: 1175: 1172: 1170: 1167: 1166: 1164: 1160: 1154: 1151: 1149: 1146: 1144: 1141: 1139: 1136: 1134: 1131: 1129: 1126: 1124: 1121: 1119: 1116: 1114: 1111: 1110: 1108: 1104: 1096: 1093: 1092: 1091: 1088: 1086: 1083: 1079: 1076: 1075: 1074: 1071: 1069: 1066: 1064: 1061: 1059: 1056: 1054: 1051: 1049: 1046: 1044: 1041: 1039: 1036: 1032: 1029: 1027: 1024: 1022: 1019: 1017: 1014: 1013: 1012: 1009: 1007: 1004: 1003: 1001: 997: 991: 988: 986: 983: 981: 978: 976: 973: 971: 968: 966: 963: 959: 956: 955: 954: 951: 949: 946: 944: 941: 939: 936: 932: 929: 928: 927: 924: 922: 919: 917: 914: 912: 909: 907: 904: 902: 899: 897: 894: 892: 889: 887: 884: 883: 881: 877: 873: 866: 861: 859: 854: 852: 847: 846: 843: 837: 836:ZPP Class ZPP 833: 832: 828: 827: 823: 819: 818: 814: 812: 811: 807: 806: 802: 800: 798: 794: 790: 786: 782: 778: 777: 772: 768: 767: 762: 757: 755: 751: 747: 743: 742: 736: 734: 730: 726: 722: 718: 714: 706: 704: 702: 698: 694: 692: 688: 684: 682: 678: 674: 669: 663: 661: 659: 655: 651: 647: 643: 639: 635: 631: 627: 623: 619: 617: 613: 609: 605: 601: 597: 593: 589: 585: 581: 576: 572: 568: 564: 560: 556: 549: 545: 541: 536: 534: 530: 525: 523: 519: 514: 512: 508: 504: 500: 496: 488: 485: 480: 477: 474: 473: 472: 471: 467: 465: 461: 457: 453: 445: 441: 437: 433: 429: 425: 421: 418: 414: 410: 406: 402: 398: 394: 393: 392: 390: 386: 382: 380: 376: 372: 364: 362: 360: 356: 352: 344: 340: 339: 338: 336: 332: 328: 324: 321:To show that 319: 317: 313: 309: 305: 301: 297: 288: 285: 281: 277: 276: 275: 273: 269: 265: 262: 258: 254: 250: 249: 244: 236: 234: 232: 228: 227: 223:. The class 222: 221: 216: 215: 210: 205: 199: 196: 193: 190: 189: 188: 186: 182: 177: 175: 171: 167: 163: 159: 155: 147: 144: 143: 142: 140: 136: 132: 128: 124: 116: 112: 108: 104: 100: 96: 91: 81: 78: 70: 60: 56: 50: 49: 43: 39: 35: 30: 21: 20: 964: 829: 815: 808: 792: 788: 784: 780: 774: 770: 764: 760: 758: 753: 749: 745: 739: 737: 732: 728: 724: 720: 716: 712: 710: 700: 696: 695: 690: 686: 685: 680: 676: 670: 667: 657: 653: 649: 645: 641: 637: 633: 632:. The class 629: 625: 624: 620: 615: 611: 607: 603: 599: 595: 591: 587: 583: 579: 574: 570: 566: 562: 558: 551: 547: 543: 539: 537: 532: 528: 526: 521: 517: 515: 510: 506: 502: 498: 494: 493:The classes 492: 483: 469: 468: 463: 462:is called a 459: 455: 451: 449: 443: 439: 435: 431: 427: 423: 416: 412: 408: 404: 400: 396: 388: 384: 383: 378: 374: 370: 369:The classes 368: 358: 354: 348: 342: 334: 330: 326: 322: 320: 315: 311: 307: 299: 295: 293: 286:algorithm B. 283: 279: 274:as follows: 267: 263: 260: 256: 252: 246: 242: 240: 224: 218: 212: 208: 206: 203: 180: 178: 169: 165: 161: 157: 153: 151: 126: 120: 73: 67:January 2013 64: 53:Please help 45: 1078:#P-complete 1016:NP-complete 931:NL-complete 450:The string 385:Definition: 337:algorithm: 59:introducing 1133:ELEMENTARY 958:P-complete 738:The class 446:) rejects. 426:is not in 419:) accepts; 407:such that 329:intersect 310:intersect 241:The class 1128:2-EXPTIME 640:contains 133:) is the 97:, co-RP, 1254:Category 1123:EXPSPACE 1118:NEXPTIME 886:DLOGTIME 803:See also 389:verifier 304:expected 1113:EXPTIME 1021:NP-hard 793:EXPTIME 776:EXPTIME 766:EXPTIME 671:ZPP is 464:witness 113:within 55:improve 1220:NSPACE 1215:DSPACE 1090:PSPACE 711:Since 470:Notes: 399:is in 343:double 270:has a 115:PSPACE 1210:NTIME 1205:DTIME 1026:co-NP 795:(see 787:, as 652:). A 646:co-RP 604:co-RP 602:(for 590:(for 544:co-RP 529:co-RP 359:co-RP 331:co-RP 312:co-RP 290:step. 284:co-RP 268:co-RP 253:co-RP 40:, or 1038:TFNP 733:coRP 731:and 721:coRP 648:and 546:and 520:and 501:and 377:and 266:and 261:both 251:and 217:and 1153:ALL 1053:QMA 1043:FNP 985:APX 980:BQP 975:BPP 965:ZPP 896:ACC 810:BPP 799:). 785:ZPP 771:ZPP 761:ZPP 754:ZPP 746:ZPP 725:ZPP 713:ZPP 701:ZPP 691:ZPP 687:ZPP 681:ZPP 677:ZPP 673:low 654:BPP 650:ZPP 638:BPP 634:BPP 630:BPP 626:ZPP 618:). 616:ZPP 548:ZPP 533:ZPP 522:ZPP 503:ZPP 466:. 422:if 395:if 379:ZPP 349:By 323:ZPP 316:ZPP 257:ZPP 243:ZPP 226:BQP 214:BPP 209:ZPP 181:ZPP 127:ZPP 121:In 103:BQP 99:BPP 1256:: 1148:RE 1138:PR 1085:IP 1073:#P 1068:PP 1063:⊕P 1058:PH 1048:AM 1011:NP 1006:UP 990:FP 970:RP 948:CC 943:SC 938:NC 926:NL 921:FL 916:RL 911:SL 901:TC 891:AC 834:: 817:RP 791:≠ 783:≠ 773:= 763:= 752:= 735:. 729:RP 723:, 719:∩ 717:RP 715:= 703:. 697:NP 693:. 689:= 683:. 679:= 644:, 642:RP 598:∈ 592:RP 586:∉ 575:V* 569:, 552:V* 542:, 540:RP 518:RP 507:NP 499:RP 497:, 495:NP 434:, 387:A 375:RP 373:, 371:NP 355:RP 335:RP 327:RP 318:. 308:RP 280:RP 264:RP 248:RP 233:. 220:RP 176:. 125:, 107:PP 105:, 101:, 95:RP 44:, 36:, 1143:R 953:P 906:L 864:e 857:t 850:v 789:P 781:P 750:P 741:P 658:X 612:X 608:x 600:X 596:x 588:X 584:x 580:X 571:w 567:x 565:( 563:V 559:x 557:( 554:w 511:f 484:x 460:w 456:V 452:w 444:w 442:, 440:x 438:( 436:V 432:w 428:X 424:x 417:w 415:, 413:x 411:( 409:V 405:w 401:X 397:x 300:k 296:k 170:n 168:( 166:p 162:n 160:( 158:p 154:n 111:P 80:) 74:( 69:) 65:( 51:.

Index

list of references
related reading
external links
inline citations
improve
introducing
Learn how and when to remove this message
Diagram of randomised complexity classes
RP
BPP
BQP
PP
P
PSPACE
complexity theory
polynomial time
complexity class
probabilistic Turing machine
Las Vegas algorithm
probabilistic Turing machine
BPP
RP
BQP
quantum computer
RP
Las Vegas algorithm
expected
Markov's Inequality
low
P

Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.