Musíte pravidelně odeslat data zpětné vazby na produkt Watson OpenScale , abyste se ujistili, že model označuje všechny změny ve vašich modelových předpovědích.

Zpětná vazba dat je nezbytná pro zachování nezaujatého modelu. Se zpětnovazební smyčkou se systém průběžně učí monitorováním účinnosti předpovědí a rekvalifikací v případě potřeby. Monitorování a využití výsledné zpětné vazby jsou jádrem strojového učení. Následující informace vám pomohou při formátování a odesílání vašich dat zpětné vazby.

Formátování dat zpětné vazby

Chcete-li správně číst data zpětné vazby, musí být správně naformátována. Služba vyhodnocení modelu přijímá následující formáty:

• Formáty souborů CSV, které lze odeslat pomocí uživatelského rozhraní nebo rozhraní REST API
• Formáty souborů JSON, které lze odeslat pouze pomocí rozhraní REST API

Tyto formáty souboru jsou definovány schématem, training_data_schema, které je dostupné v podrobnostech odběru. Chcete-li zobrazit training_data_schema, spusťte následující příkaz pomocí rozhraní API Python :

details= wos_client.subscriptions.get(subscription_id).result.to_dict()
details["entity"]["asset_properties"]["training_data_schema"]

CSV, formát

Obvykle se pro soubor CSV poskytuje data ve sloupcích a řádcích s řádkem pro názvy sloupců.

Žádné dvojité uvozovky (") jsou zapotřebí, jsou-li názvy sloupců velkými písmeny, což z nich dělá nerozlišování velkých a malých písmen pro Db2. U jiných databází a v instanci, kde jsou názvy sloupců smíšenými velkými a malými písmeny, se však velikost písmen musí shodovat.

Očekává se, že soubor CSV zpětné vazby bude mít všechny hodnoty funkcí a ručně přiřazenou hodnotu target/label. Například data školení o modelu léku obsahují hodnoty funkcí AGE, SEX, BP, CHOLESTEROL,NA,Ka cílovou hodnotu/jmenovku DRUG. Soubor CSV se zpětnou vazbou musí obsahovat hodnoty pro tato pole; příklad by vypadal jako [43, M, HIGH, NORMAL, 0.6345, 1.4587, DrugX]. Je-li pro soubor CSV se zpětnou vazbou zadáno záhlaví, jsou názvy polí mapovány pomocí záhlaví. Jinak musí být pořadí polí MUST stejné jako ve schématu školení. Další informace o školeních dat najdete v tématu Proč potřebuje produkt Watson OpenScale přístup k datům školení?

Všimněte si, že typy předpovědí vrácené vaším modelem a jmenovka/cílový sloupec ve vašich datech zpětné vazby si musí odpovídat.

Velikost souboru je momentálně omezena na 8 MB.

Pokud soubor obsahuje názvy sloupců, sloupce nemusí nutně odpovídat pořadí tabulky, ale pokud soubor neobsahuje žádné názvy sloupců, musíte odpovídat pořadí tabulky. Je možné mít sloupce, které nejsou v původních údajích o školení. Tyto sloupce jsou během zpracování ignorovány. Následující ukázka zobrazuje soubor formátu CSV s valivými naformátováním, kde uvozovky (") se používají pro názvy sloupců:

"CheckingStatus","LoanDuration","CreditHistory","LoanPurpose","LoanAmount","ExistingSavings","EmploymentDuration","InstallmentPercent","Sex","OthersOnLoan","CurrentResidenceDuration","OwnsProperty","Age","InstallmentPlans","Housing","ExistingCreditsCount","Job","Dependents","Telephone","ForeignWorker","Risk"
no_checking,28,outstanding_credit,appliances,5990,500_to_1000,greater_7,5,male,co-applicant,3,car_other,55,none,free,2,skilled,2,yes,yes,Risk
greater_200,22,all_credits_paid_back,car_used,3376,less_100,less_1,3,female,none,2,car_other,32,none,own,1,skilled,1,none,yes,No Risk
no_checking,39,credits_paid_to_date,vacation,6434,unknown,greater_7,5,male,none,4,car_other,39,none,own,2,skilled,2,yes,yes,Risk
0_to_200,20,credits_paid_to_date,furniture,2442,less_100,unemployed,3,female,none,1,real_estate,42,none,own,1,skilled,1,none,yes,No Risk
greater_200,4,all_credits_paid_back,education,4206,less_100,unemployed,1,female,none,3,savings_insurance,27,none,own,1,management_self-employed,1,none,yes,No Risk
greater_200,23,credits_paid_to_date,car_used,2963,greater_1000,greater_7,4,male,none,4,car_other,46,none,own,2,skilled,1,none,yes,Risk
no_checking,31,prior_payments_delayed,vacation,2673,500_to_1000,1_to_4,3,male,none,2,real_estate,35,stores,rent,1,skilled,2,none,yes,Risk
no_checking,37,prior_payments_delayed,other,6971,500_to_1000,1_to_4,3,male,none,3,savings_insurance,54,none,own,2,skilled,1,yes,yes,Risk
no_checking,14,all_credits_paid_back,car_new,1525,500_to_1000,4_to_7,3,male,none,4,real_estate,33,none,own,1,skilled,1,none,yes,No Risk
less_0,10,prior_payments_delayed,furniture,4037,less_100,4_to_7,3,male,none,3,savings_insurance,31,none,rent,1,skilled,1,none,yes,Risk
0_to_200,28,credits_paid_to_date,retraining,1152,less_100,less_1,2,female,none,2,savings_insurance,20,stores,own,1,skilled,1,none,yes,No Risk
less_0,17,credits_paid_to_date,car_new,1880,less_100,less_1,3,female,co-applicant,2,savings_insurance,41,none,own,1,skilled,1,none,yes,No Risk
0_to_200,39,prior_payments_delayed,appliances,5685,100_to_500,1_to_4,4,female,none,2,unknown,37,none,own,2,skilled,1,yes,yes,Risk
no_checking,32,prior_payments_delayed,radio_tv,5105,500_to_1000,1_to_4,4,male,none,5,savings_insurance,44,none,own,2,management_self-employed,1,none,yes,Risk
no_checking,38,prior_payments_delayed,appliances,4990,500_to_1000,greater_7,4,male,none,4,car_other,50,bank,own,2,unemployed,2,yes,yes,Risk
less_0,17,credits_paid_to_date,furniture,1017,less_100,less_1,2,female,none,1,car_other,30,none,own,1,skilled,1,none,yes,No Risk
less_0,33,all_credits_paid_back,car_new,3618,500_to_1000,4_to_7,2,male,none,3,unknown,31,stores,own,2,unskilled,1,none,yes,No Risk
less_0,12,no_credits,car_new,3037,less_100,less_1,1,female,none,2,car_other,31,stores,own,1,skilled,1,none,yes,No Risk
no_checking,23,prior_payments_delayed,furniture,1440,100_to_500,1_to_4,3,female,none,3,real_estate,39,stores,own,1,unskilled,1,yes,yes,No Risk
less_0,18,prior_payments_delayed,retraining,4032,less_100,1_to_4,2,female,none,2,car_other,36,none,rent,1,skilled,1,none,yes,No Risk

Formát JSON

Formát JSON se skládá z kolekce objektů s poli odpovídajícími názvům sloupců. Následující ukázka zobrazuje úplný, validně naformátovaný soubor s formátem JSON:

{ "data":
[
["less_0",10,"all_credits_paid_back","car_new",250,"500_to_1000","4_to_7",3,"male","none",2,"real_estate",23,"none","rent",1,"skilled",1,"none","yes","No Risk"],
["no_checking",23,"prior_payments_delayed","appliances",6964,"100_to_500","4_to_7",4,"female","none",3,"car_other",39,"none","own",1,"skilled",1,"none","yes","Risk"],
["0_to_200",30,"outstanding_credit","appliances",3464,"100_to_500","greater_7",3,"male","guarantor",4,"savings_insurance",51,"stores","free",1,"skilled",1,"yes","yes","Risk"],
["no_checking",23,"outstanding_credit","car_used",2681,"500_to_1000","greater_7",4,"male","none",3,"car_other",33,"stores","free",1,"unskilled",1,"yes","yes","No Risk"],
["0_to_200",18,"prior_payments_delayed","furniture",1673,"less_100","1_to_4",2,"male","none",3,"car_other",30,"none","own",2,"skilled",1,"none","yes","Risk"],
["no_checking",44,"outstanding_credit","radio_tv",3476,"unknown","greater_7",4,"male","co-applicant",4,"unknown",60,"none","free",2,"skilled",2,"yes","yes","Risk"],
["less_0",8,"no_credits","education",803,"less_100","unemployed",1,"male","none",1,"savings_insurance",19,"stores","rent",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",7,"all_credits_paid_back","car_new",250,"less_100","unemployed",1,"male","none",1,"real_estate",19,"stores","rent",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",33,"credits_paid_to_date","radio_tv",3548,"100_to_500","1_to_4",3,"male","none",4,"car_other",28,"none","own",2,"skilled",1,"yes","yes","Risk"],
["no_checking",24,"prior_payments_delayed","retraining",4158,"100_to_500","greater_7",3,"female","none",2,"savings_insurance",35,"stores","own",1,"unskilled",2,"none","yes","Risk"],
["no_checking",30,"prior_payments_delayed","appliances",5796,"unknown","4_to_7",4,"male","none",4,"car_other",49,"none","own",2,"management_self-employed",2,"none","yes","No Risk"],
["0_to_200",6,"prior_payments_delayed","furniture",2079,"500_to_1000","less_1",3,"female","none",3,"savings_insurance",35,"none","rent",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",16,"credits_paid_to_date","car_new",4172,"less_100","4_to_7",3,"male","none",3,"savings_insurance",30,"none","own",1,"skilled",1,"yes","yes","No Risk"],
["no_checking",31,"outstanding_credit","appliances",9149,"unknown","greater_7",5,"male","co-applicant",3,"unknown",61,"none","own",3,"skilled",2,"yes","yes","Risk"],
["no_checking",28,"outstanding_credit","appliances",7653,"unknown","4_to_7",4,"male","none",5,"car_other",42,"none","own",2,"skilled",1,"none","yes","No Risk"],
["0_to_200",14,"credits_paid_to_date","furniture",2056,"less_100","4_to_7",3,"male","none",3,"real_estate",20,"none","rent",1,"skilled",1,"none","yes","No Risk"],
["no_checking",29,"prior_payments_delayed","appliances",4321,"greater_1000","4_to_7",4,"male","none",4,"savings_insurance",39,"none","free",2,"management_self-employed",1,"yes","yes","Risk"],
["no_checking",8,"prior_payments_delayed","car_new",3332,"greater_1000","greater_7",3,"male","none",2,"savings_insurance",43,"none","free",2,"skilled",1,"yes","yes","Risk"],
["0_to_200",28,"credits_paid_to_date","furniture",3021,"less_100","4_to_7",3,"male","none",2,"savings_insurance",32,"none","own",2,"skilled",1,"none","yes","No Risk"],
["0_to_200",4,"prior_payments_delayed","car_new",250,"less_100","4_to_7",2,"male","none",2,"car_other",52,"bank","rent",1,"unemployed",1,"yes","yes","No Risk"],
["0_to_200",24,"outstanding_credit","business",5245,"100_to_500","greater_7",4,"male","none",4,"unknown",40,"none","own",2,"management_self-employed",1,"none","yes","Risk"],
["less_0",4,"all_credits_paid_back","radio_tv",1104,"less_100","1_to_4",2,"male","none",1,"savings_insurance",21,"none","rent",1,"skilled",1,"none","yes","No Risk"],
["less_0",10,"prior_payments_delayed","car_new",250,"less_100","4_to_7",3,"male","none",2,"real_estate",26,"none","own",1,"skilled",1,"none","yes","No Risk"],
["less_0",6,"all_credits_paid_back","radio_tv",250,"less_100","less_1",1,"female","none",2,"real_estate",19,"none","own",1,"skilled",1,"none","yes","No Risk"],
["no_checking",40,"outstanding_credit","education",7277,"unknown","greater_7",4,"male","co-applicant",5,"car_other",49,"none","own",2,"skilled",2,"yes","yes","Risk"],
["less_0",23,"credits_paid_to_date","repairs",1348,"500_to_1000","1_to_4",3,"male","none",2,"savings_insurance",35,"bank","own",2,"unskilled",1,"yes","yes","No Risk"],
["no_checking",31,"prior_payments_delayed","car_used",5011,"100_to_500","1_to_4",3,"male","none",3,"savings_insurance",41,"stores","own",1,"unskilled",1,"yes","no","No Risk"],
["0_to_200",21,"credits_paid_to_date","car_new",1658,"less_100","less_1",2,"male","none",3,"savings_insurance",27,"none","own",2,"skilled",1,"yes","yes","Risk"],
["no_checking",43,"prior_payments_delayed","appliances",6744,"greater_1000","4_to_7",3,"male","co-applicant",5,"savings_insurance",35,"none","own",2,"skilled",1,"none","yes","No Risk"],
["less_0",4,"all_credits_paid_back","car_used",250,"less_100","less_1",1,"female","none",2,"real_estate",22,"none","rent",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",24,"credits_paid_to_date","car_used",3294,"less_100","less_1",3,"male","none",4,"real_estate",25,"none","own",1,"management_self-employed",1,"none","no","No Risk"],
["greater_200",17,"credits_paid_to_date","car_new",3581,"less_100","1_to_4",2,"male","none",4,"savings_insurance",25,"none","own",1,"skilled",1,"yes","yes","No Risk"],
["no_checking",52,"outstanding_credit","appliances",9249,"unknown","greater_7",5,"male","co-applicant",4,"unknown",52,"none","free",3,"skilled",2,"yes","yes","Risk"],
["0_to_200",15,"prior_payments_delayed","retraining",398,"less_100","less_1",2,"male","co-applicant",2,"savings_insurance",44,"none","own",1,"skilled",1,"yes","yes","No Risk"],
["0_to_200",26,"credits_paid_to_date","car_used",3650,"less_100","1_to_4",3,"male","none",4,"savings_insurance",32,"stores","own",1,"unskilled",1,"none","yes","Risk"],
["less_0",4,"credits_paid_to_date","car_new",250,"100_to_500","less_1",1,"female","none",1,"real_estate",19,"bank","rent",1,"unskilled",1,"none","yes","No Risk"],
["0_to_200",16,"credits_paid_to_date","furniture",1694,"100_to_500","4_to_7",3,"female","none",3,"savings_insurance",37,"none","own",2,"skilled",1,"none","yes","No Risk"],
["no_checking",48,"outstanding_credit","education",10197,"500_to_1000","4_to_7",4,"male","co-applicant",3,"unknown",36,"none","free",2,"skilled",1,"yes","yes","Risk"],
["no_checking",43,"prior_payments_delayed","repairs",5648,"unknown","greater_7",4,"male","co-applicant",5,"unknown",49,"none","own",3,"skilled",2,"yes","yes","Risk"],
["0_to_200",13,"prior_payments_delayed","radio_tv",2217,"greater_1000","1_to_4",3,"male","co-applicant",4,"unknown",46,"bank","own",1,"unemployed",1,"yes","no","Risk"],
["greater_200",24,"credits_paid_to_date","car_used",6428,"less_100","4_to_7",4,"female","none",4,"car_other",54,"none","own",1,"skilled",1,"yes","yes","Risk"],
["0_to_200",12,"no_credits","car_new",250,"less_100","less_1",2,"female","none",1,"real_estate",20,"bank","rent",1,"unskilled",1,"none","yes","No Risk"],
["0_to_200",24,"outstanding_credit","radio_tv",9282,"500_to_1000","1_to_4",4,"male","none",5,"car_other",32,"none","own",2,"management_self-employed",1,"none","yes","No Risk"],
["less_0",8,"no_credits","radio_tv",250,"less_100","unemployed",1,"male","none",1,"real_estate",19,"bank","rent",1,"unemployed",1,"none","yes","No Risk"],
["no_checking",25,"prior_payments_delayed","radio_tv",5159,"100_to_500","1_to_4",3,"male","none",4,"car_other",53,"none","own",1,"management_self-employed",2,"yes","yes","Risk"],
["less_0",4,"credits_paid_to_date","car_new",250,"less_100","unemployed",1,"male","co-applicant",2,"real_estate",31,"stores","own",1,"skilled",1,"none","yes","No Risk"],
["no_checking",25,"outstanding_credit","business",6725,"500_to_1000","4_to_7",3,"male","none",4,"unknown",50,"none","own",2,"skilled",2,"yes","yes","Risk"],
["no_checking",19,"outstanding_credit","car_used",5092,"500_to_1000","4_to_7",4,"male","none",5,"savings_insurance",38,"none","own",1,"management_self-employed",1,"yes","yes","No Risk"],
["0_to_200",26,"all_credits_paid_back","furniture",3635,"less_100","4_to_7",4,"male","co-applicant",2,"car_other",35,"none","free",2,"skilled",1,"yes","yes","Risk"],
["0_to_200",30,"credits_paid_to_date","radio_tv",1638,"500_to_1000","1_to_4",2,"female","none",1,"savings_insurance",21,"stores","rent",1,"skilled",1,"none","yes","Risk"],
["no_checking",30,"outstanding_credit","furniture",7499,"unknown","greater_7",5,"male","co-applicant",5,"car_other",60,"none","free",2,"skilled",2,"yes","yes","Risk"],
["greater_200",5,"credits_paid_to_date","car_new",1674,"less_100","less_1",3,"female","co-applicant",3,"real_estate",31,"none","own",1,"skilled",1,"yes","yes","No Risk"],
["less_0",16,"credits_paid_to_date","car_new",250,"less_100","4_to_7",2,"male","none",1,"real_estate",29,"stores","free",1,"unskilled",1,"none","yes","No Risk"],
["less_0",32,"outstanding_credit","appliances",3511,"100_to_500","4_to_7",4,"male","none",2,"car_other",24,"none","own",1,"skilled",1,"yes","yes","No Risk"],
["0_to_200",24,"outstanding_credit","appliances",4511,"100_to_500","greater_7",4,"male","none",4,"unknown",46,"none","own",1,"skilled",1,"none","no","No Risk"],
["less_0",4,"prior_payments_delayed","car_used",1394,"less_100","1_to_4",3,"female","none",1,"savings_insurance",28,"none","own",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",23,"credits_paid_to_date","radio_tv",2795,"less_100","1_to_4",3,"male","none",3,"car_other",29,"none","own",1,"skilled",1,"none","yes","No Risk"],
["no_checking",17,"credits_paid_to_date","retraining",4331,"less_100","less_1",3,"male","none",3,"savings_insurance",32,"none","free",1,"management_self-employed",1,"none","yes","No Risk"],
["0_to_200",16,"prior_payments_delayed","radio_tv",2373,"less_100","1_to_4",3,"female","none",1,"savings_insurance",32,"none","free",2,"skilled",1,"none","yes","No Risk"],
["no_checking",38,"prior_payments_delayed","furniture",8742,"500_to_1000","4_to_7",3,"male","co-applicant",4,"unknown",46,"none","free",2,"skilled",1,"yes","yes","Risk"],
["greater_200",28,"prior_payments_delayed","radio_tv",3936,"100_to_500","1_to_4",4,"male","none",4,"car_other",36,"none","free",2,"management_self-employed",1,"yes","yes","No Risk"],
["less_0",24,"prior_payments_delayed","furniture",5098,"less_100","4_to_7",4,"female","none",2,"car_other",39,"none","own",2,"skilled",1,"none","yes","Risk"],
["no_checking",27,"prior_payments_delayed","furniture",5250,"100_to_500","1_to_4",4,"male","none",2,"car_other",53,"none","own",2,"skilled",1,"yes","yes","Risk"],
["less_0",4,"all_credits_paid_back","car_used",250,"less_100","1_to_4",1,"female","none",2,"real_estate",25,"stores","rent",1,"skilled",1,"none","yes","No Risk"],
["less_0",14,"prior_payments_delayed","vacation",4398,"less_100","1_to_4",3,"male","none",3,"car_other",31,"none","own",2,"skilled",1,"none","yes","No Risk"],
["0_to_200",4,"all_credits_paid_back","car_new",250,"less_100","1_to_4",2,"female","none",1,"real_estate",42,"none","own",1,"skilled",1,"none","yes","No Risk"],
["less_0",16,"credits_paid_to_date","furniture",2291,"less_100","less_1",3,"male","none",2,"savings_insurance",32,"stores","rent",1,"unskilled",1,"none","yes","No Risk"],
["no_checking",31,"prior_payments_delayed","furniture",7079,"100_to_500","less_1",4,"female","none",3,"car_other",43,"none","own",2,"skilled",2,"yes","yes","No Risk"],
["0_to_200",14,"credits_paid_to_date","car_new",4366,"100_to_500","1_to_4",3,"female","none",1,"car_other",37,"none","own",1,"skilled",1,"none","yes","No Risk"],
["no_checking",17,"credits_paid_to_date","repairs",3418,"100_to_500","less_1",2,"male","none",4,"car_other",43,"stores","own",2,"unskilled",1,"none","yes","Risk"],
["less_0",20,"credits_paid_to_date","furniture",902,"less_100","4_to_7",2,"male","none",1,"savings_insurance",26,"stores","own",1,"unskilled",1,"none","no","No Risk"],
["less_0",4,"credits_paid_to_date","car_used",250,"500_to_1000","less_1",1,"female","none",1,"real_estate",19,"stores","own",1,"unskilled",1,"none","yes","No Risk"],
["0_to_200",17,"no_credits","car_new",1662,"less_100","less_1",2,"female","none",2,"savings_insurance",32,"none","own",2,"skilled",1,"yes","yes","No Risk"],
["0_to_200",27,"credits_paid_to_date","education",1272,"greater_1000","1_to_4",3,"male","none",2,"savings_insurance",37,"stores","own",2,"unskilled",1,"none","no","No Risk"],
["less_0",14,"prior_payments_delayed","radio_tv",1640,"500_to_1000","less_1",2,"male","none",2,"real_estate",29,"none","rent",2,"skilled",1,"none","yes","No Risk"],
["greater_200",20,"credits_paid_to_date","car_new",250,"100_to_500","1_to_4",2,"male","none",3,"savings_insurance",44,"none","own",1,"skilled",1,"yes","yes","No Risk"],
["no_checking",33,"all_credits_paid_back","education",6126,"less_100","4_to_7",5,"male","none",3,"savings_insurance",33,"stores","rent",1,"unskilled",1,"none","yes","Risk"],
["0_to_200",8,"prior_payments_delayed","car_new",731,"greater_1000","1_to_4",3,"female","none",2,"savings_insurance",33,"bank","rent",1,"unemployed",2,"none","yes","No Risk"],
["no_checking",31,"prior_payments_delayed","vacation",5086,"100_to_500","greater_7",5,"male","none",3,"savings_insurance",44,"none","free",2,"management_self-employed",1,"none","yes","Risk"],
["less_0",15,"prior_payments_delayed","furniture",3986,"less_100","less_1",2,"male","none",2,"car_other",19,"bank","rent",1,"unskilled",1,"none","yes","No Risk"],
["0_to_200",13,"credits_paid_to_date","car_new",1700,"less_100","4_to_7",3,"female","none",2,"car_other",39,"none","own",1,"skilled",1,"none","yes","Risk"],
["0_to_200",4,"all_credits_paid_back","car_new",250,"less_100","1_to_4",2,"female","none",1,"real_estate",29,"none","own",1,"skilled",1,"none","yes","No Risk"],
["less_0",16,"prior_payments_delayed","car_new",2056,"less_100","4_to_7",2,"male","none",2,"unknown",34,"stores","own",1,"skilled",1,"none","yes","No Risk"],
["no_checking",32,"prior_payments_delayed","car_used",5145,"500_to_1000","4_to_7",4,"male","none",4,"savings_insurance",40,"none","own",1,"skilled",1,"none","yes","No Risk"],
["greater_200",24,"outstanding_credit","business",8570,"100_to_500","4_to_7",4,"male","none",4,"unknown",47,"none","own",2,"skilled",1,"yes","yes","Risk"],
["no_checking",16,"all_credits_paid_back","vacation",2928,"less_100","1_to_4",2,"female","none",1,"real_estate",29,"none","own",1,"management_self-employed",1,"none","yes","No Risk"],
["less_0",25,"credits_paid_to_date","car_new",2049,"less_100","less_1",2,"male","none",1,"real_estate",33,"none","own",1,"skilled",1,"none","yes","No Risk"],
["no_checking",24,"prior_payments_delayed","car_used",3838,"500_to_1000","1_to_4",4,"female","none",3,"savings_insurance",41,"none","own",2,"skilled",1,"none","yes","No Risk"],
["0_to_200",21,"all_credits_paid_back","car_new",1653,"100_to_500","less_1",1,"female","none",1,"car_other",33,"bank","rent",1,"unemployed",1,"none","yes","No Risk"],
["less_0",7,"credits_paid_to_date","car_new",250,"less_100","1_to_4",1,"female","none",2,"savings_insurance",19,"stores","rent",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",20,"credits_paid_to_date","retraining",2553,"less_100","1_to_4",3,"female","none",3,"savings_insurance",27,"stores","rent",1,"unskilled",2,"none","yes","No Risk"],
["less_0",39,"prior_payments_delayed","appliances",820,"100_to_500","4_to_7",2,"male","none",1,"savings_insurance",26,"stores","own",1,"skilled",1,"none","yes","No Risk"],
["0_to_200",20,"credits_paid_to_date","retraining",2810,"less_100","1_to_4",2,"female","none",3,"car_other",38,"bank","rent",2,"unemployed",1,"yes","yes","No Risk"],
["greater_200",4,"prior_payments_delayed","furniture",3262,"less_100","1_to_4",2,"male","none",3,"savings_insurance",31,"none","free",1,"management_self-employed",1,"yes","yes","No Risk"],
["less_0",28,"outstanding_credit","business",2891,"100_to_500","4_to_7",4,"male","co-applicant",2,"savings_insurance",40,"none","own",1,"skilled",1,"yes","yes","Risk"],
["0_to_200",28,"credits_paid_to_date","education",1531,"less_100","4_to_7",2,"male","none",2,"car_other",28,"stores","own",1,"skilled",1,"yes","yes","Risk"],
["no_checking",41,"prior_payments_delayed","repairs",7507,"unknown","4_to_7",5,"male","co-applicant",5,"car_other",41,"none","own",2,"skilled",2,"none","yes","No Risk"],
["0_to_200",4,"prior_payments_delayed","vacation",250,"500_to_1000","4_to_7",2,"male","none",3,"savings_insurance",27,"none","own",1,"skilled",1,"none","yes","No Risk"]
]
}

Přenos dat zpětné vazby

Data zpětné vazby můžete odeslat ze souboru CSV přímo do uživatelského rozhraní produktu Watson OpenScale . Chcete-li odeslat soubor JSON, můžete použít produkt Watson Studio.

Odeslání souboru CSV

Chcete-li odeslat soubor CSV, použijte volbu Přidat data zpětné vazby. Chcete-li v rámci výukového programu postupovat podle následujících pokynů, otevřete a zkopírujte obsah souboru credit_feedback_data.csv .

1. Na panelu dashboard Watson OpenScale klepněte na dlaždici implementace.
2. V okně implementace modelu klepněte na volbu Monitory konfigurace .
3. V navigačním panelu klepněte na volbu Kvalita.
4. Klepněte na položku Zpětná vazba a poté klepněte na volbu Přidat data zpětné vazby.

5. Vyberte soubor CSV, který obsahuje data zpětné vazby, a klepněte na tlačítko Otevřít. Pro výukový program vyberte soubor credit_feedback_data.csv , který jste stáhli.

   Velikost souboru je momentálně omezena na 8 MB.

6. V rozevírací nabídce klepněte na oddělovač polí a poté klepněte na tlačítko Vybrat.

Přidání souboru CSV poskytuje data zpětné vazby do vašeho modelu.

Odeslání souboru JSON

1. Spusťte produkt Watson Studio a přejděte do projektu, který obsahuje daný model.
2. Stáhněte soubor JSON.

3. Na kartě Implementace vašeho projektu Watson Studio klepněte na odkaz model , klepněte na kartu Test a vyberte ikonu vstupu JSON.

   

4. Nyní otevřete stažený soubor JSON a zkopírujte obsah do pole JSON na kartě Test . Klepnutím na tlačítko Předpověz odešlete a zabodujte informační obsah školení do svého modelu.

Zobrazení výsledků

Chcete-li výsledek zkontrolovat okamžitě, na stránce Insights vyberte implementaci, klepněte na jednu z metrik Kvalita a poté klepněte na volbu Vyhodnotit kvalitu nyní.

Další kroky

Konfigurace vyhodnocení koncového bodu

Nadřízené téma: Správa dat pro vyhodnocení modelu v produktu Watson OpenScale