/*************************************************************************************** * * WRITEPAD(r): Handwriting Recognition Engine (HWRE) and components. * Copyright (c) 2001-2016 PhatWare (r) Corp. All rights reserved. * * Licensing and other inquires: * Developer: Stan Miasnikov, et al. (c) PhatWare Corp. * * WRITEPAD HWRE is free software: you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation, either version 3 of the License, or * (at your option) any later version. * * THE MATERIAL EMBODIED ON THIS SOFTWARE IS PROVIDED TO YOU "AS-IS" * AND WITHOUT WARRANTY OF ANY KIND, EXPRESS, IMPLIED OR OTHERWISE, * INCLUDING WITHOUT LIMITATION, ANY WARRANTY OF MERCHANTABILITY OR * FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL PHATWARE CORP. * BE LIABLE TO YOU OR ANYONE ELSE FOR ANY DIRECT, SPECIAL, INCIDENTAL, * INDIRECT OR CONSEQUENTIAL DAMAGES OF ANY KIND, OR ANY DAMAGES WHATSOEVER, * INCLUDING WITHOUT LIMITATION, LOSS OF PROFIT, LOSS OF USE, SAVINGS * OR REVENUE, OR THE CLAIMS OF THIRD PARTIES, WHETHER OR NOT PHATWARE CORP. * HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH LOSS, HOWEVER CAUSED AND ON * ANY THEORY OF LIABILITY, ARISING OUT OF OR IN CONNECTION WITH THE * POSSESSION, USE OR PERFORMANCE OF THIS SOFTWARE. * See the GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with WritePad. If not, see . * **************************************************************************************/ #ifndef MLP_HEADER_INCLUDED #define MLP_HEADER_INCLUDED #include "snn.h" #if !MLP_PRELOAD_MODE || MLP_LEARN_MODE #include #endif // -------------------- Defines ------------------------------------------------ #define MLP_EXPTABL_MAX 8 #define MLP_EXPTABL_SIZE 512 // -------------------- Defines ------------------------------------------------ #define MLP_UNROLL_CYCLES 1 #define MLP_ROLLING_LEARN 0 #define MLP_F_CELL_CMASK 0x0001 #define MLP_F_CELL_PMASK 0x0002 // ----------------------------------------------------------------------------- #define MLP_ABS(x) (((x) > 0) ? (x):(-(x))) #define MLP_ERR(e) ((e)*(e)*(e)) //#define MLP_ERR(e) (((e) > 0) ? (e)*(e) : -((e)*(e))) #if MLP_UPSCALE #if MLP_BYTE_NET #define MLP_MAX_INT_C 127 #define MLP_MAX_INT_S 255 #if MLP_INT_MODE #define x_UPCO_C << 3 #define x_DNCO_C >> 3 #define x_UPCO_S << 8 #define x_DNCO_S >> 8 #define x_DNDNCO_S >> 16 #else #define x_UPCO_C *(8) #define x_DNCO_C *(0.125) #define x_UPCO_S *(256) #define x_DNCO_S *(0.00390625) #define x_DNDNCO_S *(1.52587890625E-05) #endif #else // SHORT mode #define MLP_MAX_INT_C 32767 #define MLP_MAX_INT_S 4095 #if MLP_INT_MODE #define x_UPCO_C << 9 #define x_DNCO_C >> 9 #define x_UPCO_S << 12 #define x_DNCO_S >> 12 #define x_DNDNCO_S >> 24 #else #define x_UPCO_C *(512) #define x_DNCO_C *(0.001953125) #define x_UPCO_S *(4096) #define x_DNCO_S *(0.000244140625) #define x_DNDNCO_S *(5.96046447753906E-08) #endif #endif // MLP_BYTE_NET #else // #define MLP_MAX_INT_C 63.9 // #define MLP_MAX_INT_C 15.875 #define MLP_MAX_INT_C 127 #define MLP_MAX_INT_S 1.0 #define x_UPCO_C #define x_DNCO_C #define x_UPCO_S #define x_DNCO_S #define x_DNDNCO_S #endif #if MLP_EMULATE_INT && !(MLP_INT_MODE) #define FINT_C_OF(x) ((fint_c)(x)) #define FINT_S_OF(x) ((fint_s)(x)) #else #define FINT_C_OF(x) ((flong)(x)) #define FINT_S_OF(x) ((flong)(x)) #endif #if MLP_INT_MODE #define FINT_C_LD(x) (((x) > -MLP_MAX_INT_C && (x) < MLP_MAX_INT_C) ? (fint_c)(x) : (((x) > 0) ? MLP_MAX_INT_C : -MLP_MAX_INT_C)) #else #define FINT_C_LD(x) (x) #endif // -------------------- Structures and classes --------------------------------- typedef struct { _USHORT inp_ind; fint_c bias; fint_c weights[MLP_CELL_MAXINPUTS]; #if MLP_LEARN_MODE // learning float err; float sbias; float psbias; float sws[MLP_CELL_MAXINPUTS]; float psws[MLP_CELL_MAXINPUTS]; _INT num_sws; _INT num_psws; _UCHAR flags; // _USHORT out_ind; _INT num_changes; float prev_val; #endif } mlp_cell_type, * p_mlp_cell_type; typedef struct { flong sum_delt; _INT num_sum; } mlp_layer_type, * p_mlp_layer_type; typedef struct { _UCHAR id_str[32]; _INT num_layers; _INT num_inputs; _INT num_outputs; fint_s exp_tabl[MLP_EXPTABL_SIZE]; mlp_cell_type cells[MLP_NET_MAX_NUMCELLS]; mlp_layer_type layers[MLP_NET_NUMLAYERS]; } mlp_net_type, * p_mlp_net_type; // ---------------- Functions -------------------------------------------------- _INT CountNetResult(p_UCHAR inps, p_UCHAR outs, p_mlp_data_type mlpd, _INT lang ); fint_s CountCellSignal(_INT nc, p_mlp_data_type mlpd); RECO_DATA p_mlp_net_type GetNetData(_INT lang); #if !MLP_PRELOAD_MODE _INT LoadNet(FILE * file, p_mlp_data_type mlpd, _INT langID ); _INT InitNet(_INT type, p_mlp_data_type mlpd, _INT langID ); #endif #if MLP_LEARN_MODE _INT FillExpTable(flong ic, p_mlp_data_type mlpd); _INT InitNetWeights(_INT type, float ic, p_mlp_data_type mlpd, _INT langID); _INT CountNetError(float * desired_outputs, float zc, p_mlp_data_type mlpd, _INT langID); _INT ModifyNetDeltas(_INT flags, p_mlp_data_type mlpd, _INT langID); _INT AdjustNetWeights(_INT flags, float * lcs, float ic, p_mlp_data_type mlpd, _INT langID); _INT ShakeNetWeights(_INT flags, float lc, p_mlp_data_type mlpd, _INT langID); _INT CountNetStats(_INT mode, p_mlp_data_type mlpd, _INT langID); _INT SaveNet(FILE *file, p_mlp_data_type mlpd, _INT langID ); _INT DumpNet(FILE *file, p_mlp_data_type mlpd, _INT langID ); _INT CalcHiddenLayerError(_INT layer_st, _INT layer_len, p_mlp_data_type mlpd); _INT CountLayerStats(_INT mode, _INT layer_num, _INT layer_st, _INT layer_len, p_mlp_data_type mlpd); _INT InitCellWeights(_INT type, float ic, _INT ncell, p_mlp_data_type mlpd); #endif // MLP_LEARN_MODE #endif // MLP_HEADER_INCLUDED