사업비검증TF 인공지능 모형

preprocessing.ipynb

- finaldata.pkl : shape (3690789, 144)

EXCC_CONF_DE

USE_DE

RECHCT_EXCC_ORG_REG_YN

TRSC_EVDC_SE_CD

TRSC_EVDC_NO

TRSC_PFMC_REG_DT

TRSC_PFMC_CNCL_DT

TRSC_PFMC_STT_CD

CNCL_FORM_SE_CD

CNCL_BEF_TRSC_PFMC_STT_CD

TRSC_DE

SPLY_AMT_DFFR_RSN_CN

VAT_HDLG_SE_CD

SPLR_BCMP_NM

SPLR_BSNSR_REG_NO

SPLR_BSTP_NM

SPLR_BUCDT_NM

SPLR_ADDR

SPLR_BSNS_REG_STT_CD

SPLR_TAXT_TP_SE_CD

CLSBS_SPLR_EXCT_RSN_CN

RECHCT_EXCT_ID

ANNL

AGRT_ORGN_ID

GVCTM_PTTN_PAY_YN

TTL_BSNS_AMT

GOV_CTRB_AMT

UPAY_GOV_CTRB_AMT

MM_TRSF_BNDS_AMT

LCGVN_ALOT_CASH_AMT

LCGVN_ALOT_SPOT_AMT

PRVT_ALOT_CASH_AMT

PRVT_ALOT_SPOT_AMT

AGRT_VLID_YN

EXCT_LMT_YN

NAT_RND_BSNS_YN

DVLM_DTRS_YN

MRS_DTRS_YN

AGRT_ORGN_ROLE_SE_CD

ORGN_TTL_BSNS_AMT

AGRT_BNDS_AMT

ODY_TRSF_BNDS_AMT

DBT_TRSF_BNDS_AMT

SPOT_CROV_AMT

PRYY_CASH_CROV_AMT

PRYY_SPOT_CROV_AMT

TEMP_MM_TRSF_BNDS_AMT

TEMP_ODY_TRSF_BNDS_AMT

TEMP_DBT_TRSF_BNDS_AMT

PRVT_ALOT_AMT_BEF_YN

AGRT_ORGN_CHNG_SE_CD

BEXP_DVS_EXCT_YN

LCGVN_PAYM_WAY_SE_CD

EXCT_LMT_AMT

CRMT_ALLW_YN

TESSR_ECPT_SE_CD

SBJT_KIND_SE_CD

SBJT_SE_CD

CCLT_TFEE_AMT

TECL_CD

HIRK_RECHCT_USE_ITEPD_YN

PTC_ITEPD_CD

DPTC_ITEPD_CD_x

CASH_SPOT_SE_CD

USE_SPLY_AMT

USE_VAT_AMT

USE_CROV_SPLY_AMT

USE_CROV_VAT_AMT

EVDC_PPS_ATCH_DOC_ID

VAT_HDLG_RSTO_AMT

RSTO_CONF_SPLY_AMT

RSTO_CONF_VAT_AMT

RSTO_CONF_CROV_SPLY_AMT

RSTO_CONF_CROV_VAT_AMT

ORGN_CNLK_DTRS_YN

ETC_EVDC_ECTN_CD

ETC_EVDC_ECTN_RSN_CN

USE_AMT_SE_CD

AST_AMT

CRAST_AMT

NCAS_AMT

DEBT_AMT

CRLBT_AMT

NCL_AMT

CPTL_TTL_SUM_AMT

CPTL_AMT

SALE_AMT

SALE_TTL_GAIN_AMT

OPRFT_AMT

ROS_AMT

CEOS_AMT

COTA_COST_DEDU_BEF_NPRFT_AMT

COTA_COST_AMT

TTM_NPRFT_AMT

EQMT_UNQ_NO

CNTR_STT_CD

EQMT_SE_CD

ACQS_DE

ACQS_AMT

CEDC_YN

EQMT_REG_AMT

REGTM_ACQS_AMT

RECHCT_USE_ITEPD_ID

RECHCT_USE_TRSC_PFMC_ID

AGORG_BZEXC_ID

DSAPVL_HDLG_STT_CD

DSAPVL_RSN_SE_CD

DSAPVL_RSN_DTL_CN

LCOST_DTL_SN

PAY_AMT

LCOST_PAY_SE_CD

LCOST_OBJT_SE_CD

PAY_AGRT_ORGN_ID

PAY_OBJT_PTCPR_ID

PAY_OBJT_USER_ID

PAY_OBJT_RCHSP_SN

TXIVC_APRB_NO

ISSU_DE

WRT_DE

SPPNR_BSNSR_REG_NO

SPPNR_BSTP_NM

SPPNR_BUCDT_NM

SPPNR_ZIP_NO

SPPNR_ADDR

SUM_AMT

ART_REG_IT

FSU_AMT

FAT_AMT

FCO_AMT

FEC_AMT

FPU_SUM_AMT

FRCS_TAXT_TP_INFO_CN

FRCS_ADDR

USE_DT

TXMT

SECH_AMT

DDC_YN_DCSN_SE_CD

CHAC_TRSF_RQS_SN

DPTC_ITEPD_CD_y

CHAC_TRSF_RQS_DTIT_SN

RQS_AMT

ART_NM

EVDC_DOC_ID

BSNSR_REG_NO

- df=dummies(df, "DSAPVL_RSN_SE_CD") :  159열로 확장

data collection.ipynb

- mergedata.pkl 을 EXCC_CONF_DE>='20160101' 해서 shape (3172234, 247)

df = pd.merge(data, eqpt, how = 'left', on="RECHCT_USE_ITEPD_ID")

df.shape = (3178356, 279)

lcost.shape = (2110515 ,27) // lcost가 뭔데?(data1~data33)

df = pd.merge(df, lcost, how = 'left', on="RECHCT_USE_ITEPD_ID")

Tdata를 templtdata.pkl로 떨굼 // Tdata가 뭔데?(T1~T22)

cdata를 templcdata.pkl 로 떨굼 // cdata가 뭔데?(C1~C44)

rdata = pd.read_excel('/home/grusls/backupdata/rawdata/R.xls')

rdata를 templrdata.pkl로 떨굼

ddata = pd.read_excel('/home/grusls/backupdata/rawdata/DD.xls')

ddata를 templddata.pkl로 떨굼

df = pd.merge(df, Tdata, how = 'left', on="RECHCT_USE_TRSC_PFMC_ID")

df = pd.merge(df, cdata, how = 'left', on="RECHCT_USE_TRSC_PFMC_ID")

orgdataorgdata = pd.read_excel('/home/grusls/backupdata/rawdata/BSNSRNO.xls')

orgdata = orgdata.rename(columns = {"ORGN_ID" : "AGRT_ORGN_ID"})

orgdata를 temporg.pkl 로 떨굼

df.shape = (3690375, 378)

df = pd.merge(df, rdata, how = 'left', on="RECHCT_USE_TRSC_PFMC_ID")

df.shape = (3690375, 396)

df = pd.merge(df, ddata, how = 'left', on="RECHCT_USE_TRSC_PFMC_ID")

df = pd.merge(df, orgdata, how = 'left', on="AGRT_ORGN_ID")

df를 finaldata.pkl로 떨굼

test notebook.ipynb

 

 

model_others.json

model_others_test.json

model_VAT.json

model_VAT_test.json