MARC record from Internet Archive

LEADER: 04606cam 2200781 a 4500
001 ocm40251635
003 OCoLC
005 20191030065136.0
008 981030s1999 enka b 001 0 eng
010 $a 98048438
040 $aDLC$beng$cDLC$dUKM$dOHX$dNLGGC$dBAKER$dBTCTA$dYDXCP$dOCLCG$dDEBBG$dEXW$dTULIB$dBDX$dOCLCF$dOCLCO$dOCLCQ$dI8M$dOCLCO$dOCLCQ$dOCLCA$dOCLCO$dOCLCA$dGZM$dOCLCQ$dBGU$dOCLCO$dAU@$dOCLCQ$dUKMGB
015 $aGB98V6592$2bnb
016 7 $a011020692$2Uk
019 $a40838977$a123071377$a1120789883
020 $a1852330953$q(alk. paper)
020 $a9781852330958$q(alk. paper)
035 $a(OCoLC)40251635$z(OCoLC)40838977$z(OCoLC)123071377$z(OCoLC)1120789883
050 00 $aQA76.87$b.H87 1999
072 7 $aQA$2lcco
082 00 $a006.3/2$221
084 $a54.72$2bcl
084 $aPN 96$2blsrissc
084 $aPN 95$2blsrissc
084 $aQH 700$2rvk
084 $aSK 835$2rvk
084 $aST 300$2rvk
084 $a15$2sdnb
084 $a27$2sdnb
084 $a28$2sdnb
084 $aDAT 717f$2stub
100 1 $aHusmeier, Dirk,$d1964-
245 10 $aNeural networks for conditional probability estimation :$bforecasting beyond point predictions /$cDirk Husmeier.
260 $aLondon ;$aNew York :$bSpringer,$c℗♭1999.
300 $axxiii, 275 pages :$billustrations ;$c24 cm.
336 $atext$btxt$2rdacontent
337 $aunmediated$bn$2rdamedia
338 $avolume$bnc$2rdacarrier
490 1 $aPerspectives in neural computing
504 $aIncludes bibliographical references (pages 267-272) and index.
505 00 $tConventional forecasting and Takens' embedding theorem --$tImplications of observational noise --$tImplications of dynamic noise --$tA Universal Approximator Network for Predicting Conditional Probability Densities --$tA single-hidden-layer network --$tAn additional hidden layer --$tRegaining the conditional probability density --$tMoments of the conditional probability density --$tInterpretation of the network parameters --$tGaussian mixture model --$tDerivative-of-sigmoid versus Gaussian mixture model --$tComparison with other approaches --$tPredicting local error bars --$tIndirect method --$tComplete kernel expansion: Conditional Density Estimation Network (CDEN) and Mixture Density Network (MDN) --$tDistorted Probability Mixture Network (DPMN) --$tMixture of Experts (ME) and Hierarchical Mixture of Experts (HME) --$tSoft histogram --$tAppendix: The moment generating function for the DSM network --$tA Maximum Likelihood Training Scheme --$tThe cost function --$tA gradient-descent training scheme --$tOutput weights --$tKernel widths --$tRemaining weights --$tInterpretation of the parameter adaptation rules --$tDeficiencies of gradient descent and their remedy --$tBenchmark Problems --$tLogistic map with intrinsic noise --$tStochastic combination of two stochastic dynamical systems --$tBrownian motion in a double-well potential --$tDemonstration of the Model Performance on the Benchmark Problems --$tLogistic map with intrinsic noise --$tMethod --$tResults --$tStochastic coupling between two stochastic dynamical systems --$tMethod --$tResults --$tAuto-pruning.
650 0 $aNeural networks (Computer science)
650 0 $aDistribution (Probability theory)$xData processing.
650 7 $aDistribution (Probability theory)$xData processing.$2fast$0(OCoLC)fst00895603
650 7 $aNeural networks (Computer science)$2fast$0(OCoLC)fst01036260
650 17 $aNeurale netwerken.$2gtt
650 17 $aWaarschijnlijkheidsverdelingen.$2gtt
650 17 $aSchattingstheorie.$2gtt
650 17 $aPrognoses.$2gtt
650 7 $aBedingte Wahrscheinlichkeitsverteilung$2gnd
650 7 $aNeuronales Netz$2gnd
650 7 $aPra diktionsanalyse$2gnd
650 7 $aBade$2gnd
650 7 $aScha tzung$2gnd
776 08 $iOnline version:$aHusmeier, Dirk, 1964-$tNeural networks for conditional probability estimation.$dLondon ; New York : Springer, ℗♭1999$w(OCoLC)654309156
830 0 $aPerspectives in neural computing.
938 $aBaker & Taylor$bBKTY$c99.00$d99.00$i1852330953$n0003232879$sactive
938 $aBrodart$bBROD$n53709357$c$89.95
938 $aBaker and Taylor$bBTCP$n98048438
938 $aOtto Harrassowitz$bHARR$nhar000660283
938 $aYBP Library Services$bYANK$n100143980
029 1 $aAU@$b000014255337
029 1 $aDEBBG$bBV012337567
029 1 $aNLGGC$b176843183
029 1 $aOHX$bhar000660283
029 1 $aYDXCP$b100143980
029 1 $aUKMGB$b011020692
994 $aZ0$bP4A
948 $hNO HOLDINGS IN P4A - 156 OTHER HOLDINGS