úÎ)"&Þ     None-Signed real value in the logarithmic domain. Positive component Negative component Smart LogSigned constructor. 3Make LogSigned from a positive, log-domain number. 3Make LogSigned from a negative, log-domain number. $Shift LogSigned to a normal domain.  Change the  to either negative    or positive  .  !"#$    !"#$None =Gradient with nonzero values stored in a logarithmic domain. ? Since values equal to zero have no impact on the update phase ? of the SGD method, it is more efficient to not to store those  components in the gradient. @Add normal-domain double to the gradient at the given position. EAdd log-domain, singed number to the gradient at the given position. 8Construct gradient from a list of (index, value) pairs. ; All values from the list are added at respective gradient  positions. EConstruct gradient from a list of (index, signed, log-domain number) C pairs. All values from the list are added at respective gradient  positions. :Collect gradient components with values in normal domain. 1Empty gradient, i.e. with all elements set to 0. 5Perform parallel unions operation on gradient list.  Experimental version. %"Parallel unions in the Par monad. & %    & %None'4Type synonym for mutable vector with Double values. Vector of parameters. !Dataset with elements of x type. 1SGD parameters controlling the learning process. Size of the batch Regularization variance Number of iterations Initial gain parameter 2After how many iterations over the entire dataset  the gain parameter is halved Default SGD parameter values. 8Pure version of the stochastic gradient descent method. ;Monadic version of the stochastic gradient descent method. : A notification function can be used to provide user with 1 information about the progress of the learning. (9Add up all gradients and store results in normal domain. )%Scale the vector by the given value. *IApply gradient to the parameters vector, that is add the first vector to  the second one. 'SGD parameter values Gradient for dataset element Dataset Starting point  SGD result SGD parameter values Notification run every update Gradient for dataset element Dataset Starting point  SGD result ()*+   '()*+,      !"#$ %&'()*+,-./012 sgd-0.2.2Numeric.SGD.LogSignedNumeric.SGD.Grad Numeric.SGD LogSignedposneg logSignedfromPosfromNegtoNorm toLogFloatGradaddaddLfromList fromLogListtoListempty parUnionsParaDatasetSgdArgs batchSizeregVariterNumgain0tausgdArgsDefaultsgdsgdMbase Data.EitherLeftlogfloat-0.12.1Data.Number.LogFloatLogFloatRightzero$fNumLogSigned$fNFDataLogSigned$fOrdLogSigned $fEqLogSigned parUnionsP insertWithMVectaddUpscaleapplysample