You are on page 1of 21
202247 KES AGA / 2022 Summer Entrance Examination Department of Creative Informatics abe Graduate School of Information Science and Technology Al) 3 iS 1 The University of Tokyo AIS : Creative Informatics 1 RRAFWRELARARA eR 1 RREMOSMET, COMMMPER OBC. LORBO TBCH SARSS MIC BRSTERAT SCE. REARO SY THREE FHEWE FIC, DOM dDEOTH LO. BERRONRS KOM, SBHSHLOMBSSESATKEATSCE PEPE LUMEN FASS RWC. INSTRUCTIONS, 1. Do not open this booklet until the start of the examination is announced. 2. Write your examinee ID number below on this cover page. 3. You may write on the back of the answer sheet. 4, Write your examinee ID number and the problem number inside the top blanks of each sheet. 5. Do not bring the answer sheet or this booklet out of this room. BES / Examinee 1D #14 nr PEOMMETS. Bia 1,2,...,.niHL, fp BMRA TEMS 1 BRS HBB L. BMD WML ay OWT f(x) = 00 LFS. Ly a =r RHF ERO FRM (01,..., cn) ERAT TM UMITNS, SHI, BABB OL file) SRA KF SRTTAMEBER, JL TLORCO ROMAINE ENS. COME WFO KS HMS NS. xt Yoke) & Py sete De et olFER ORR, (1) Bim 1,2,...,m FROMM IL, BR di(a) = fi(a) — fila — 1) BERL, dia) a KL CHMMNCS 3 EET So (P) KATORMI TY AL Ag BA FB. Step 0: #=1,2,...,.nKHL. 7 O0E TS. Step 1: AF OPE r BART: i= 1,2, NDB dj(aj +1) ERAKTS HMOMEIE 7 EL. 2, 2,412 93. ROBVICBR Le (-l) r=5,n=3, UT fi, fa, fa BAFOMELSEDETS, TORK di, do,ds ld HMM CHS CEE L. MRT VAV AA Ag LORE. (1-2) (af, 23,-..,0) ERATURE TS. CHA (P) OMMBTHS ILE, UFOR PERMIT CLMAIATH SCL ERE, ip diet +2) Sin ea?) (1-3) RAT VAY AL Ag #8 (P) OBMRSHWNT SOLER (2) (1) OFEMIMED READ Me UBA CIA, RAT VAY KL A to (P) ORG MEMNTSLARSRV. TOT, SVMF SRBC, (P)On er eEN 2NN €(1,2,...,n} E RE {0,1,...,7} KMERALROME (PH) EBA So x BAG SaAe Fe q i PD) ae le a aEFEROBB, = 1,...,N COMMODORE oy(R) LIT. ROMWIEBA Le (1) N > 20K EROPABM ATS gwalo) & flo} OBBAOT. an(R) Re. 2) (P) DBE ga(r) EMF SDO, WAV MET VY AAO BE 1S AACA, WR, LOFTY LAB Ap EMR. (2:8) BIMOEHERT ATY XA Ay (KD (P) DBMS NSC YER. (2-4) SHORT VY XA Ap OHRME BRT VAY) AL Ac OMBMEBR Ke PERE LS fy Sn DH HO FINE RE Le Problem 1 Let n and r be positive integers. For i = 1,2,...,7, let fi be a univariate real-valued function defined in the integer domain and let fi(2) be —oo for negative integer 2, Any non-negative integer solution (a1,..-,tq) that satisfies 7, 2; = r is called & feasible solution. In addition, a feasible solution that maximizes the objective function Shur Jil) is called an optimal solution and the objective function value at the solution is called the optimal value. This problem is expressed as follows. Maximize Aled )) subject to > ay is a non-negative integer, i= 1,...,2 (1) For i= 1,2,...,n and non-negative integer a, define the function 4)(a) := fi(a] - fifa —1) and assume that di(q) is non-increasing in terms of a. Apply the following greedy algorithm Ag to (P) Step 0: For i= 1,2,...,n, set aj <0. Step 1: Repeat the following procedure for r times: Let y be any index i that maximizes d(e;-+ 1) among i= 1,2,...,1, and set ay @ ty +1 Answer the following questions. (1-1) Let r = 5,n=3, and let fa, fo, fs take the following values. Notice that di, d2, ds, are non-incroasing. Answer the solution obtained by the greedy algorithm Ac. fila) fhla)}-2 1 -14 -40 fola)| 0-8 12 -22 (1-2) Let (xf, 7},...,2%) be 0 feasible solution. Show that it is an optimal solution of (P) if and only if the following condition holds. max | di(of +1) < min | di(a7) Sime aa (1-3) Show that the greedy algorithm Ag outputs an optimal solution of (P). (2) Unless the non-increasing assumption of (1) holds, the greedy algorithm Ag does not always output an optimal solution of (P). To apply dynamic programming, we consider the following problem (Pf) in which n and r in (P) are replaced with NV € {1,2,....n} and Re {0,1,...,r}, respectively. y Maximize VA a (Pa) subject to 7; is a non-negative integer, i iN ‘The optimal value of the problem is denoted by gv(R). Answer the following questions. (2-1) Express gy(R) only with gr—s(¢) and fry(c) for any non-negative integer c in the case of N > 2. (2-2) Write a pscudo-code of @ dynamic programming algorithm within 15 lines to output the optimal value ga(r) of (P). Hereafter, this algorithm is called Ap. (2-3) Show that the optimal value of (P) is obtained by the dynamic programming algorithm Ap. (2-4) Answer the computational complexity of the dynamic programming algorithm. Ap and the computational complexity of the greedy algorithm Ag. Ignore the computational cost of caloulating fi,.... fa 20224F EAB / 2022 summer Entrance Examination Department of Creative Informatics aewnaeex Graduate School of Information : Science and Technology All eR 2 The University of Tokyo ANI Creative Informatics 2 ERAS HSELERARA ERR 1 RRVAOSMET, COMEMFEMABT Ee. 2 LORMLN FBICH SRRSS MI BASSE BAT SCE 3. MSRMOSOCHKBR HAVER, 5b bHcdeoTo LD. 4 RERLOMESHKMAC, RRBSH LOMESSERAT KEATS CE 5, PERRSLOUMERFGESIBSETE. INSTRUCTIONS 1. Donot open this booklet until the start of the examination is announced, 2. Write your examinee ID number below on this cover page. 3. You may write on the back of the answer sheet. 4, Write your examinee 1D number and the problem number inside the top blanks of each sheet. 5, Do not bring the answer sheet or this booklet out of this room. EE /EramineeID ‘This page is blank. LORY. ‘This page is blank. 38 21 BRR LEY Ea —AM CONT y PRE HR S, BY, PAT Y RI = EDD PRA) — BAOHRICKE RT 7A VOT 7 FO-K ATS. Oy SE OMULBS LU 2 OHMS DW, FHAP EO IAT YJ Fi DIRS BIBT 7 KADF— FI y PRD SHES DOME RENT y MAOF-FOPE 1 Ly HULO Ey PRY BRET BIBER 0 Osa K

You might also like