id sid tid token lemma pos 5d86nz8378d 1 1 deep deep ADJ 5d86nz8378d 1 2 neural neural ADJ 5d86nz8378d 1 3 networks network NOUN 5d86nz8378d 1 4 have have AUX 5d86nz8378d 1 5 recently recently ADV 5d86nz8378d 1 6 gained gain VERB 5d86nz8378d 1 7 immense immense ADJ 5d86nz8378d 1 8 popularity popularity NOUN 5d86nz8378d 1 9 , , PUNCT 5d86nz8378d 1 10 especially especially ADV 5d86nz8378d 1 11 for for ADP 5d86nz8378d 1 12 developing develop VERB 5d86nz8378d 1 13 surrogate surrogate ADJ 5d86nz8378d 1 14 models model NOUN 5d86nz8378d 1 15 for for ADP 5d86nz8378d 1 16 applications application NOUN 5d86nz8378d 1 17 in in ADP 5d86nz8378d 1 18 the the DET 5d86nz8378d 1 19 field field NOUN 5d86nz8378d 1 20 of of ADP 5d86nz8378d 1 21 science science NOUN 5d86nz8378d 1 22 and and CCONJ 5d86nz8378d 1 23 engineering engineering NOUN 5d86nz8378d 1 24 . . PUNCT 5d86nz8378d 2 1 these these DET 5d86nz8378d 2 2 surrogate surrogate ADJ 5d86nz8378d 2 3 models model NOUN 5d86nz8378d 2 4 are be AUX 5d86nz8378d 2 5 computationally computationally ADV 5d86nz8378d 2 6 inexpensive inexpensive ADJ 5d86nz8378d 2 7 and and CCONJ 5d86nz8378d 2 8 approximate approximate VERB 5d86nz8378d 2 9 the the DET 5d86nz8378d 2 10 input input NOUN 5d86nz8378d 2 11 - - PUNCT 5d86nz8378d 2 12 output output NOUN 5d86nz8378d 2 13 mapping mapping NOUN 5d86nz8378d 2 14 of of ADP 5d86nz8378d 2 15 the the DET 5d86nz8378d 2 16 expensive expensive ADJ 5d86nz8378d 2 17 physical physical ADJ 5d86nz8378d 2 18 simulators simulator NOUN 5d86nz8378d 2 19 . . PUNCT 5d86nz8378d 3 1 surrogates surrogate NOUN 5d86nz8378d 3 2 can can AUX 5d86nz8378d 3 3 be be AUX 5d86nz8378d 3 4 used use VERB 5d86nz8378d 3 5 efficiently efficiently ADV 5d86nz8378d 3 6 for for ADP 5d86nz8378d 3 7 problems problem NOUN 5d86nz8378d 3 8 that that PRON 5d86nz8378d 3 9 require require VERB 5d86nz8378d 3 10 expensive expensive ADJ 5d86nz8378d 3 11 , , PUNCT 5d86nz8378d 3 12 repetitive repetitive ADJ 5d86nz8378d 3 13 simulations simulation NOUN 5d86nz8378d 3 14 such such ADJ 5d86nz8378d 3 15 as as ADP 5d86nz8378d 3 16 inverse inverse ADJ 5d86nz8378d 3 17 modeling modeling NOUN 5d86nz8378d 3 18 , , PUNCT 5d86nz8378d 3 19 uncertainty uncertainty NOUN 5d86nz8378d 3 20 quantification quantification NOUN 5d86nz8378d 3 21 , , PUNCT 5d86nz8378d 3 22 and and CCONJ 5d86nz8378d 3 23 propagation propagation NOUN 5d86nz8378d 3 24 or or CCONJ 5d86nz8378d 3 25 design design NOUN 5d86nz8378d 3 26 of of ADP 5d86nz8378d 3 27 experiments experiment NOUN 5d86nz8378d 3 28 tasks task NOUN 5d86nz8378d 3 29 . . PUNCT 5d86nz8378d 4 1 this this DET 5d86nz8378d 4 2 work work NOUN 5d86nz8378d 4 3 focuses focus VERB 5d86nz8378d 4 4 on on ADP 5d86nz8378d 4 5 investigating investigate VERB 5d86nz8378d 4 6 deep deep ADJ 5d86nz8378d 4 7 learning learning NOUN 5d86nz8378d 4 8 approaches approach NOUN 5d86nz8378d 4 9 for for ADP 5d86nz8378d 4 10 the the DET 5d86nz8378d 4 11 inverse inverse ADJ 5d86nz8378d 4 12 and and CCONJ 5d86nz8378d 4 13 forward forward ADJ 5d86nz8378d 4 14 solutions solution NOUN 5d86nz8378d 4 15 of of ADP 5d86nz8378d 4 16 physical physical ADJ 5d86nz8378d 4 17 systems system NOUN 5d86nz8378d 4 18 . . PUNCT 5d86nz8378d 5 1 for for ADP 5d86nz8378d 5 2 the the DET 5d86nz8378d 5 3 first first ADJ 5d86nz8378d 5 4 part part NOUN 5d86nz8378d 5 5 , , PUNCT 5d86nz8378d 5 6 a a DET 5d86nz8378d 5 7 inverse inverse ADJ 5d86nz8378d 5 8 model model NOUN 5d86nz8378d 5 9 is be AUX 5d86nz8378d 5 10 developed develop VERB 5d86nz8378d 5 11 that that SCONJ 5d86nz8378d 5 12 maps map NOUN 5d86nz8378d 5 13 noisy noisy ADJ 5d86nz8378d 5 14 and and CCONJ 5d86nz8378d 5 15 gappy gappy ADJ 5d86nz8378d 5 16 observations observation NOUN 5d86nz8378d 5 17 to to ADP 5d86nz8378d 5 18 a a DET 5d86nz8378d 5 19 high high ADV 5d86nz8378d 5 20 - - PUNCT 5d86nz8378d 5 21 dimensional dimensional ADJ 5d86nz8378d 5 22 input input NOUN 5d86nz8378d 5 23 . . PUNCT 5d86nz8378d 6 1 this this DET 5d86nz8378d 6 2 problem problem NOUN 5d86nz8378d 6 3 is be AUX 5d86nz8378d 6 4 ill ill ADV 5d86nz8378d 6 5 - - PUNCT 5d86nz8378d 6 6 posed pose VERB 5d86nz8378d 6 7 since since SCONJ 5d86nz8378d 6 8 one one PRON 5d86nz8378d 6 9 may may AUX 5d86nz8378d 6 10 not not PART 5d86nz8378d 6 11 be be AUX 5d86nz8378d 6 12 able able ADJ 5d86nz8378d 6 13 to to PART 5d86nz8378d 6 14 uniquely uniquely ADV 5d86nz8378d 6 15 recover recover VERB 5d86nz8378d 6 16 the the DET 5d86nz8378d 6 17 high high ADV 5d86nz8378d 6 18 - - PUNCT 5d86nz8378d 6 19 dimensional dimensional ADJ 5d86nz8378d 6 20 input input NOUN 5d86nz8378d 6 21 given give VERB 5d86nz8378d 6 22 the the DET 5d86nz8378d 6 23 noisy noisy ADJ 5d86nz8378d 6 24 and and CCONJ 5d86nz8378d 6 25 sparse sparse ADJ 5d86nz8378d 6 26 output output NOUN 5d86nz8378d 6 27 . . PUNCT 5d86nz8378d 7 1 traditional traditional ADJ 5d86nz8378d 7 2 inverse inverse ADJ 5d86nz8378d 7 3 modeling modeling NOUN 5d86nz8378d 7 4 techniques technique NOUN 5d86nz8378d 7 5 require require VERB 5d86nz8378d 7 6 a a DET 5d86nz8378d 7 7 large large ADJ 5d86nz8378d 7 8 number number NOUN 5d86nz8378d 7 9 of of ADP 5d86nz8378d 7 10 simulator simulator NOUN 5d86nz8378d 7 11 runs run NOUN 5d86nz8378d 7 12 , , PUNCT 5d86nz8378d 7 13 which which PRON 5d86nz8378d 7 14 is be AUX 5d86nz8378d 7 15 computationally computationally ADV 5d86nz8378d 7 16 expensive expensive ADJ 5d86nz8378d 7 17 . . PUNCT 5d86nz8378d 8 1 to to PART 5d86nz8378d 8 2 address address VERB 5d86nz8378d 8 3 this this DET 5d86nz8378d 8 4 issue issue NOUN 5d86nz8378d 8 5 and and CCONJ 5d86nz8378d 8 6 solve solve VERB 5d86nz8378d 8 7 the the DET 5d86nz8378d 8 8 inverse inverse ADJ 5d86nz8378d 8 9 problem problem NOUN 5d86nz8378d 8 10 efficiently efficiently ADV 5d86nz8378d 8 11 , , PUNCT 5d86nz8378d 8 12 we we PRON 5d86nz8378d 8 13 develop develop VERB 5d86nz8378d 8 14 an an DET 5d86nz8378d 8 15 inverse inverse ADJ 5d86nz8378d 8 16 surrogate surrogate ADJ 5d86nz8378d 8 17 model model NOUN 5d86nz8378d 8 18 that that PRON 5d86nz8378d 8 19 hinges hinge VERB 5d86nz8378d 8 20 on on ADP 5d86nz8378d 8 21 the the DET 5d86nz8378d 8 22 concepts concept NOUN 5d86nz8378d 8 23 of of ADP 5d86nz8378d 8 24 deep deep ADJ 5d86nz8378d 8 25 generative generative ADJ 5d86nz8378d 8 26 machine machine NOUN 5d86nz8378d 8 27 learning learning NOUN 5d86nz8378d 8 28 and and CCONJ 5d86nz8378d 8 29 is be AUX 5d86nz8378d 8 30 trained train VERB 5d86nz8378d 8 31 with with ADP 5d86nz8378d 8 32 limited limited ADJ 5d86nz8378d 8 33 data datum NOUN 5d86nz8378d 8 34 in in ADP 5d86nz8378d 8 35 a a DET 5d86nz8378d 8 36 supervised supervised ADJ 5d86nz8378d 8 37 setting setting NOUN 5d86nz8378d 8 38 . . PUNCT 5d86nz8378d 9 1 the the DET 5d86nz8378d 9 2 other other ADJ 5d86nz8378d 9 3 part part NOUN 5d86nz8378d 9 4 of of ADP 5d86nz8378d 9 5 this this DET 5d86nz8378d 9 6 work work NOUN 5d86nz8378d 9 7 develops develop VERB 5d86nz8378d 9 8 an an DET 5d86nz8378d 9 9 efficient efficient ADJ 5d86nz8378d 9 10 forward forward ADV 5d86nz8378d 9 11 surrogate surrogate ADJ 5d86nz8378d 9 12 model model NOUN 5d86nz8378d 9 13 that that PRON 5d86nz8378d 9 14 is be AUX 5d86nz8378d 9 15 trained train VERB 5d86nz8378d 9 16 with with ADP 5d86nz8378d 9 17 limited limited ADJ 5d86nz8378d 9 18 data datum NOUN 5d86nz8378d 9 19 for for ADP 5d86nz8378d 9 20 the the DET 5d86nz8378d 9 21 forward forward ADJ 5d86nz8378d 9 22 solution solution NOUN 5d86nz8378d 9 23 . . PUNCT 5d86nz8378d 10 1 here here ADV 5d86nz8378d 10 2 , , PUNCT 5d86nz8378d 10 3 we we PRON 5d86nz8378d 10 4 develop develop VERB 5d86nz8378d 10 5 a a DET 5d86nz8378d 10 6 data data NOUN 5d86nz8378d 10 7 - - PUNCT 5d86nz8378d 10 8 driven drive VERB 5d86nz8378d 10 9 model model NOUN 5d86nz8378d 10 10 with with ADP 5d86nz8378d 10 11 a a DET 5d86nz8378d 10 12 hybrid hybrid ADJ 5d86nz8378d 10 13 framework framework NOUN 5d86nz8378d 10 14 by by ADP 5d86nz8378d 10 15 combining combine VERB 5d86nz8378d 10 16 multiscale multiscale ADJ 5d86nz8378d 10 17 methods method NOUN 5d86nz8378d 10 18 and and CCONJ 5d86nz8378d 10 19 the the DET 5d86nz8378d 10 20 deep deep ADJ 5d86nz8378d 10 21 neural neural ADJ 5d86nz8378d 10 22 networks network NOUN 5d86nz8378d 10 23 for for ADP 5d86nz8378d 10 24 stochastic stochastic ADJ 5d86nz8378d 10 25 multiscale multiscale ADJ 5d86nz8378d 10 26 pde pde NOUN 5d86nz8378d 10 27 systems system NOUN 5d86nz8378d 10 28 , , PUNCT 5d86nz8378d 10 29 in in ADP 5d86nz8378d 10 30 particular particular ADJ 5d86nz8378d 10 31 , , PUNCT 5d86nz8378d 10 32 flow flow VERB 5d86nz8378d 10 33 in in ADP 5d86nz8378d 10 34 porous porous ADJ 5d86nz8378d 10 35 media medium NOUN 5d86nz8378d 10 36 . . PUNCT 5d86nz8378d 11 1 we we PRON 5d86nz8378d 11 2 then then ADV 5d86nz8378d 11 3 introduce introduce VERB 5d86nz8378d 11 4 a a DET 5d86nz8378d 11 5 bayesian bayesian ADJ 5d86nz8378d 11 6 framework framework NOUN 5d86nz8378d 11 7 to to ADP 5d86nz8378d 11 8 this this DET 5d86nz8378d 11 9 hybrid hybrid ADJ 5d86nz8378d 11 10 model model NOUN 5d86nz8378d 11 11 to to PART 5d86nz8378d 11 12 perform perform VERB 5d86nz8378d 11 13 uncertainty uncertainty NOUN 5d86nz8378d 11 14 quantification quantification NOUN 5d86nz8378d 11 15 and and CCONJ 5d86nz8378d 11 16 propagation propagation NOUN 5d86nz8378d 11 17 tasks task NOUN 5d86nz8378d 11 18 . . PUNCT 5d86nz8378d 12 1 overall overall ADJ 5d86nz8378d 12 2 , , PUNCT 5d86nz8378d 12 3 deep deep ADJ 5d86nz8378d 12 4 learning learning NOUN 5d86nz8378d 12 5 models model NOUN 5d86nz8378d 12 6 developed develop VERB 5d86nz8378d 12 7 in in ADP 5d86nz8378d 12 8 this this DET 5d86nz8378d 12 9 work work NOUN 5d86nz8378d 12 10 offer offer VERB 5d86nz8378d 12 11 a a DET 5d86nz8378d 12 12 significant significant ADJ 5d86nz8378d 12 13 acceleration acceleration NOUN 5d86nz8378d 12 14 in in ADP 5d86nz8378d 12 15 terms term NOUN 5d86nz8378d 12 16 of of ADP 5d86nz8378d 12 17 computational computational ADJ 5d86nz8378d 12 18 cost cost NOUN 5d86nz8378d 12 19 and and CCONJ 5d86nz8378d 12 20 accuracy accuracy NOUN 5d86nz8378d 12 21 for for ADP 5d86nz8378d 12 22 both both CCONJ 5d86nz8378d 12 23 the the DET 5d86nz8378d 12 24 inverse inverse ADJ 5d86nz8378d 12 25 and and CCONJ 5d86nz8378d 12 26 forward forward ADJ 5d86nz8378d 12 27 solutions solution NOUN 5d86nz8378d 12 28 of of ADP 5d86nz8378d 12 29 physical physical ADJ 5d86nz8378d 12 30 systems system NOUN 5d86nz8378d 12 31 . . PUNCT