iMed User Manual
Selelekela
1.1. Morero
Sepheo sa sena web kopo ke ho nka tlhahisoleseding e tala le ho lumella ho e fetola ka mokhoa o fanang ka liphello tse molemo ha ho etsoa liqeto. Sena e ka ba ho koetlisa mohlala ka data e tala kapa ho bolela esale pele sephetho ho sebelisa mehlala le tlhahlobo.
1.2. Navigational menu
Lenane la ho sesa le kaholimo ho leqephe le tšoere lihokelo tsohle ho fihla moo o hlokang ho ba teng. Haeba u ka lahleha, u ka tobetsa motsu o ka morao ho fihla leqepheng leo u le tloaetseng, u khutlele hae, kapa u fumane leqephe leo u le batlang ka har'a menyetla ea ho tsamaea.
1.3. Akhaonto
Haeba ha u na ak'haonte, u tlameha ho ingolisa ho sebelisa sesebelisoa. Ho etsa joalo, tobetsa konopo ea ak'haonte e kaholimo ho le letona ebe o tobetsa regista. Ebe u kenya lebitso la hau la mosebelisi, password, le lengolo-tsoibila ho tsoela pele.
Haeba u se u ntse u e-na le ak'haonte, kena ka lebitso la hau le password.
Leqephe la Lehae
Ka ho tobetsa lintho tse ka letsohong le letšehali la leqephe, tlhaloso ea e 'ngoe le e 'ngoe e tla hlaha bohareng ba leqephe ho u thusa ho utloisisa seo e 'ngoe le e 'ngoe e se etsang.
iMedBot
Sesebelisoa sa iMedBot se hlahisa sebopeho se khothalletsang tšebelisano e bonolo ea basebelisi le baemeli, se nolofalletsang ho bolela esale pele le koetliso ea mohlala. E sebetsa e le mohato oa pele oa ho fetola liphetho tsa lipatlisiso tse tebileng tsa ho ithuta ho ba sesebelisoa sa inthanete, se nang le monyetla oa ho tsosa merero ea lipatlisiso sebakeng sena. Buka ea eona e fapaneng ea mosebelisi e ka fumanoa mona.
Tlhahlobo ea Lintlha
4.1. Khutlisa Li-subsets
Karolo ena e lumella mosebelisi ho hlophisa pokello ea data ea bona. U ka khetha ho kenya dataset e ncha kapa ho sebelisa e seng e ntse e le teng ho tsoa ho menu e theoha.
Hang ha dataset e kentsoe, u ka khetha hore na u batla ho nka khato efe ka ho tobetsa e 'ngoe ea likhetho tse ka lehlakoreng le letšehali la menyu.
4.1.1. Khutlisa li-subsets tse ipapisitseng le Lisefe
Karolo ena e u lumella ho fumana karolo e nyane ea pokello ea data ea mantlha ho ipapisitsoe le lihloela tse fanoeng. Khetha litekanyetso tseo u li batlang karolong e nyenyane ebe u khetha likholomo tseo u batlang hore li bontšoe ho dataset ea ho qetela.
4.1.2. Khutlisa Liphetho tse Hlophisitsoeng
Sena se khutlisa dataset ka mokhoa o hlophisitsoeng. Kgetha kholomo e shebilweng, tatellano ya ho hlopha, palo ya mela e tla kgutla, le hore na ke dikholomo dife tse tla bontshwa sephethong sa ho qetela.
4.1.3. Eketsa Dataset
Sena se lumella mosebedisi ho atolosa kholomo e le nngwe e bolokilweng e le bukantswe ho tafola ya nnete eo mosebedisi a ka e fetolang. E nka dataset e behiloeng 'me e tsamaisa se hlokoang ke mosebelisi ho ea karolong e kaholimo ho fetisisa. Taba ea pele, kenya setlankane sa data se kenyeletsang kholomo e nang le dataset e hlophisitsoeng. Haeba kholomo e hlokang ho atolosoa e fumaneha ka bo eona, khetha hore na u ka atolosa kholomo efe le hore na u ntšoa litšitiso life. Tobetsa romela 'me u ka khona view lintlha tsa hau e le likholomo tsa tafole ho fapana le data e behiloeng.
4.2. Kopanya Files
Ka ho khetha le ho kenya li-dataset tse ngata ka ho tobetsa ctrl (command for mac), sena se tla li kopanya hore e be datha e le 'ngoe e kholoanyane ho feta ho sebelisetsoa ho hong.
Khetha feela li-dataset tsohle 'me u tlatse lintlha tse hlokahalang. Sena se tla boloka dataset e ncha ho sesebelisoa sa iMed mme e tla fumaneha bakeng sa ho jarolleloa.
4.3. Mesebetsi ea Plot
Karolo ena e lumella mosebelisi ho rala datha ea bona. Khetha e 'ngoe ea likhetho ho menu e ka letsohong le letšehali ebe u tlatsa libaka tse hlokahalang ho fumana morero oa hau. Ka tlase ke mefuta ea merero eo u ka e etsang ho tsoa ho data ea hau:
4.4. Tlhahlobo ea Lipalo-palo
Karolo ena e re lumella ho etsa liteko tsa lipalo ho dataset ea rona. Khetha teko eo u tla e sebelisa ho tsoa ho menu e ka letsohong le letšehali 'me u tlatse likarolo ho etsa liteko. Ka tlase ke mefuta ea liteko tse fumanehang:
ODPAC
5.1. Ithute
Leqephe lena le kenyelletsa tlhaloso e khuts'oane ea mofuta o mong le o mong oa lisebelisoa tse fumanehang leqepheng lena. Ho tobetsa konopo e kaholimo ho karolo ka 'ngoe ho tla hokahanya le leqephe le leng ho lumella mosebelisi ho sebelisa kapa ho ithuta haholoanyane ka sehlooho.
5.1.1. Epistasis
Leqephe lena le re lumella ho sebelisa MBS, mokhoa oa ho batla ho ithuta ho tsoa ho data. Ka ho khetheha, e re lumella ho ithuta epistasis, tšebelisano pakeng tsa liphatsa tsa lefutso tse peli kapa ho feta tse amang phenotype. Sena se na le thuso ho profile mafu karolong ea liphatsa tsa lefutso. Mekhoa e tloaelehileng ha e tšoanelehe ho sebetsana le lintlha tse phahameng tse fumanoang lithutong tsa genome-wide association (GWAS). Multiple Beam Search (MBS) algorithm e lumella ho lemoha liphatsa tsa lefutso tse sebelisanang ka lebelo le potlakileng haholo. Kenya data eo u batlang ho e sebelisa ebe u kenya likarolo tse hlokahalang. Bakeng sa tlhaiso-leseling e batsi, fumana pampiri e felletseng mona.
5.1.2. Lintho tsa Kotsi
Leqephe lena le re lumella ho sebelisa sephutheloana sa IGain ho ithuta litšebelisano lipakeng tsa data. E ithuta ka ho khetheha litšebelisano ho tsoa ho data ea boemo bo holimo e sebelisa lipatlisiso tsa heuristic. Mokhoa ona o theha mokhoa oa Exhaustive_IGain o hlahisitsoeng pele ho ithuta litšebelisano ho tsoa ho data e tlase. Kenya data ebe u kenya likarolo tse hlokahalang. Lintlha tse ling mabapi le litekanyo tsa IS le iGain li ka fumanoa Mona.
5.1.3. Mehlala ea Boprofeta
Karolo ena e lumella tšebeliso ea mefuta ea ho bolela esale pele e seng e hahiloe ka holim'a mefuta ea ho ithuta ea mochini ho potlakisa ts'ebeliso ea eona. Sena se lumella ts'ebeliso ea bona ntle le ts'ebeliso ea likhoutu le boiphihlelo ba pele ho bolela esale pele mefuta ba sebelisa dataset ea bona. Ho na le mefuta e mengata ea ho bolela esale pele e fumanehang ho mosebelisi ho kenyelletsa Logistic, Regression, Support Vector Machines (SVMs), Lifate tsa Qeto, le tse ling tse ngata. Lethathamo le felletseng la mekhoa ea ho bolela esale pele le fumanoa ka lehlakoreng le letona la leqephe mona.
5.2. Boprofeta
Karolo ena e lumella likhakanyo ho tsoa ho mofuta o arolelanoeng o kentsoeng pele. Qala ka ho kenya mofuta o arolelanoang haeba o sa etsoa joalo. Ebe u khetha mohlala oo u ka o sebelisang ho bolela esale pele ka ho tobetsa lebitso la mohlala. Ebe u kenya data bakeng sa mofuta oa ho bolela esale pele hore o e sebelise. Sena se ka etsoa ka letsoho ho sebelisa foromo e ka tlase ho leqephe kapa ho sebelisa template e fumanehang bakeng sa ho jarolleloa. Haeba u sebelisa template, kenya dataset file ebe o tobetsa ho romella ho amohela ponelopele ea mohlala.
5.3. Tšehetso ea Qeto
Tšehetso ea liqeto e fana ka lihlopha 'me e ka tataisa khetho ea phekolo ho tsoa boitsebisong bo fanoeng tsamaisong. E koetlisitsoe ho tsoa ho data ho khothaletsa mokhoa o nepahetseng oa kalafo o ipapisitseng le likarolo tsa mokuli. Lintlha tse ling mabapi le Clinical Decision Support Systems (CDSS) li ka fumanoa mona.
Keletso ea Sistimi e nka likarolo tsa mokuli mme e khothaletsa mokhoa oa kalafo le ho bolela esale pele monyetla oa nako e tlang oa metastasis ea lilemo tse 5. Ts'ebetso ea Basebelisi e nka likarolo tsa mokuli le mokhoa oa kalafo ho bolela esale pele monyetla oa nako e tlang oa metastasis ea lilemo tse 5 ho latela kalafo ea hajoale ho fapana le kalafo e nepahetseng.
MBIL
Markov Blanket le Interactive Risk Factor Learner (MBIL) ke algorithm e ithutang mabaka a kotsi a le mong le a kopanetsoeng a nang le tšusumetso e tobileng sephethong sa mokuli. Tobetsa "ea ho MBIL" ho fetisetsoa ho Python Package Index (PyPI) bakeng sa sephutheloana sa MBIL se fumanehang mona. Lintlha tse ling mabapi le MBIL li ka fumanoa ho BMC Bioinformatics.
Lisebelisoa tsa data
Karolo ena e lumella mosebelisi ho bona le ho kenya li-dataset tse ncha ho web kopo.
7.1. Sheba Li-database Tsohle li Teng
Ho bona li-datasets tsohle li fumaneha, tobetsa feela "Show Available Datasets".
7.2. Kenya Dataset
Ho kenya dataset, tobetsa "Share Your Datasets" ebe u tlatsa lintlha tse hlokahalang joalokaha ho boletsoe ho webleqephe. Ntlha ea pele, kenya dataset 'me u tlatse likarolo tse hlokahalang.
Ebe, tlatsa likarolo tse ka tlase kapa kenya mongolo file ka litaba tse tlatsitsoeng. An exampLe ea mokhoa oa ho hlophisa tlhahisoleseling e le hore kopo e ka e utloisisa e fanoe ka tlase.
Mehlala
Karolo ena e lumella mosebelisi ho bona mefuta e fumanehang ho bona le ho arolelana mohlala.
8.1. Sheba Mefuta Eohle e Fumanehang
Ho bona mefuta eohle e fumanehang, tobetsa "Show Available Models".
8.2. Arolelana Mohlala
Ho arolelana mohlala, tobetsa ho "Arolelana Mehlala ea Hao" ebe u kenya mohlala file koetlisitsoeng ke tensor flow kapa PyTorch.
8.2.1. Dataset e Amanang
Joale o lokela ho kenya dataset e amanang le eona e kenyelletsang lihlooho. Sehlopha/leibole ea pokello ea lintlha e lokela ho ba kholomong ea ho qetela.
8.2.2. Li-predictors le lintlha tsa Sehlopha
Haeba dataset e kenyelletsa likarolo tsohle, foromo ea sebopeho e ka tloloa ka mor'a ho kenya dataset. Leha ho le joalo, haeba li sa kenyelletsoa kaofela, tlhahisoleseding ena e tlameha ho fanoa tlhalosong file kapa ka har'a sebopeho sa sebopeho. Khetha khetho ho tloha ho theoha ho bonts'a hore na u ikemiselitse ho fana ka li-predictors le lintlha tsa sehlopha joang.
Haeba u sebelisa khetho ea tlhaloso, u ka tlatsa likarolo kapa ua kenya mongolo file ka litaba tse tlatsitsoeng. An example ea mokhoa oa ho hlophisa tlhahisoleseding e fanoe ka tlase.
Litokomane / Lisebelisoa
 |
Lisebelisoa tsa iMed Web Kopo [pdf] Bukana ea Mosebelisi iMed, iMed Web Kopo, Web Kopo |