SG11201407897YA - Latent variable model estimation apparatus and method - Google Patents

Abstract

(i2) mwu ft iz s-5 iv x m £ titz s issaj n d9) mm IHMMi _ (41) SI5S4>!H B 2013^12^ 5 0(05.12.2013) W HRO I P CT (10) WO 2013/179579 A1 (51) G06N 5/04 { 2006.01) (21) PCT/JP2013/002900 (22) H|®t±J®IS: 2013^4^ 30 0(30.04.2013) (25) (26) @(8&M©Wl§: 0*f§ (30) fiBfcfif '— 1 £: 61/653,855 2012^5^ 31 0(31.05.2012) US (71) tiili® A:_0 CORPORATION) [JP/JP]; T 1088001 E 2S T I7f1^ Tokyo (JP). (72) % BJ # : H # if ¥ (FUJIMAKI, Ryohei); T 1088001 7f 1 ^0$lt Tokyo (JP). IS TK. ®g(MORINAGA, Satoshi); T 1088001 SET |7f1^ Tokyo (JP). == (74) ftHA: #}it, ^(IWAKABK, Iuyuki cl al ); = T 1040031 M 51 #5 rp £ E j? v ~ T g 8 # 7 ^ Tokyo (JP). (54) Title: HIDDEN-VARIABLE-MODEL ESTIMATION DEVICE AND METHOD = (54) |: ®*l£»^ (57) Abstract: Provided is a hidden-variable-model estima­ tion device that makes it possible to implement high-speed model selection even if the number of candidate models in­ creases exponentially with the number of hidden states and the number of types of measurement probabilities. A vari- ation-probability calculation unit (71) calculates a variation probability by maximizing a reference value defined as the lower bound of the Laplace approximation of the estimated value of marginalized log-likelihood function with a respect to a complete variable. A model estimation unit (72) pro­ duces an estimation of the optimal hidden-variable model by estimating the measurement-probability type of and paramet­ ers for each hidden state. A convergence determination unit (73) determines whether or not the reference value that the variation-probability calculation unit (71) used when calcu­ lating the variation probability converged. (57) mm : as £ t 1 it. I- -So 2IA, #®*ittfi8lc*JLTt& jlJfS^CDffi— £ Ittfk'Mt' -S £®*l£»^7 ;U£tlt^?\"3o i 4XM£SP7 3 * (8i) Jt^ia (^ro&i^PBy, IS ft nJ f b): AE, AG, AL, AM, AO, AT, AU, AZ, BA, BB, BG, BH, BN, BR, BW, BY, BZ, CA, CH, CL, CN, CO, CR, CU, CZ, DE, DK, DM, DO, DZ, EC, EE, EG, ES, FI, GB, GD, GE, GH, GM, GT, HN, HR, HU, ID, IL, IN, IS, JP, KE, KG, KM, KN, KP, KR, KZ, LA, LC, LK, LR, LS, LT, LU, LY, MA, MD, ME, MG, MK, MN, MW, MX, MY, MZ, NA, NG, NI, NO, NZ, OM, PA, PE, PG, PH, PL, PT, QA, RO, RS, RU, RW, SC, SD, SE, SG, SK, SL, SM, ST, SV, SY, TH, TJ, TM, TN, TR, TT, TZ, UA, UG, US, UZ, VC, VN, ZA, ZM, ZW. (84) > ±T(Dmm(Dfcm% 11^ nlfb): ARIPO (BW, GH, GM, KE, LR, LS, MW, MZ, NA, RW, SD, SL, SZ, TZ, UG, ZM, ZW), 3. — =j V T (AM, AZ, BY, KG, KZ, RU, TJ, TM), 3 — • V / ^ (AL, AT, BE, BG, CH, CY, CZ, DE, DK, EE, ES, FI, FR, GB, GR, HR, HU, IE, IS, IT, LT, LU, LV, MC, MK, MT, NL, NO, PL, PT, RO, RS, SE, SI, SK, SM, TR), OAPI (BF, BJ, CF, CG, CI, CM, GA, GN, GQ, GW, ML, MR, NE, SN, TD, TG). - (&&£ 21 &(3)) = r-> \ f § \ f 0\ ir, o\ i> i-H cn 71 w 72 w 73 w 71 72 73 VARIATION-PROBABILITY CALCULATION UNIT MODEL ESTIMATION UNIT CONVERGENCE DETERMINATION UNIT O CJ