{"id":"https://openalex.org/W7165131345","doi":"https://doi.org/10.1007/s11334-026-00642-8","title":"From diagnosis to repair: A model-driven framework for root cause analysis of machine learning pipelines","display_name":"From diagnosis to repair: A model-driven framework for root cause analysis of machine learning pipelines","publication_year":2026,"publication_date":"2026-06-19","ids":{"openalex":"https://openalex.org/W7165131345","doi":"https://doi.org/10.1007/s11334-026-00642-8"},"language":"en","primary_location":{"id":"doi:10.1007/s11334-026-00642-8","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11334-026-00642-8","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11334-026-00642-8.pdf","source":{"id":"https://openalex.org/S151239915","display_name":"Innovations in Systems and Software Engineering","issn_l":"1614-5046","issn":["1614-5046","1614-5054"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Innovations in Systems and Software Engineering","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://link.springer.com/content/pdf/10.1007/s11334-026-00642-8.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5060873383","display_name":"Emmanuel Charleson Dapaah","orcid":null},"institutions":[{"id":"https://openalex.org/I74656192","display_name":"University of G\u00f6ttingen","ror":"https://ror.org/01y9bpm73","country_code":"DE","type":"education","lineage":["https://openalex.org/I74656192"]}],"countries":["DE"],"is_corresponding":true,"raw_author_name":"Emmanuel Charleson Dapaah","raw_affiliation_strings":["Institute of Computer Science, University of Goettingen, Goldschmidtstra\u00dfe 7, Goettingen, 37077, Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Computer Science, University of Goettingen, Goldschmidtstra\u00dfe 7, Goettingen, 37077, Germany","institution_ids":["https://openalex.org/I74656192"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5078674128","display_name":"Jens Grabowski","orcid":"https://orcid.org/0000-0003-2994-3531"},"institutions":[{"id":"https://openalex.org/I74656192","display_name":"University of G\u00f6ttingen","ror":"https://ror.org/01y9bpm73","country_code":"DE","type":"education","lineage":["https://openalex.org/I74656192"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Jens Grabowski","raw_affiliation_strings":["Institute of Computer Science, University of Goettingen, Goldschmidtstra\u00dfe 7, Goettingen, 37077, Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Computer Science, University of Goettingen, Goldschmidtstra\u00dfe 7, Goettingen, 37077, Germany","institution_ids":["https://openalex.org/I74656192"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5060873383"],"corresponding_institution_ids":["https://openalex.org/I74656192"],"apc_list":{"value":2190,"currency":"EUR","value_usd":2790},"apc_paid":{"value":2190,"currency":"EUR","value_usd":2790},"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.94813703,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":null,"biblio":{"volume":"22","issue":"2","first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.38089999556541443,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.38089999556541443,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10260","display_name":"Software Engineering Research","score":0.1014999970793724,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.08460000157356262,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/hyperparameter","display_name":"Hyperparameter","score":0.765999972820282},{"id":"https://openalex.org/keywords/boosting","display_name":"Boosting (machine learning)","score":0.59579998254776},{"id":"https://openalex.org/keywords/debugging","display_name":"Debugging","score":0.579200029373169},{"id":"https://openalex.org/keywords/pipeline-transport","display_name":"Pipeline transport","score":0.5561000108718872},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.5275999903678894},{"id":"https://openalex.org/keywords/perceptron","display_name":"Perceptron","score":0.508899986743927},{"id":"https://openalex.org/keywords/gradient-boosting","display_name":"Gradient boosting","score":0.48249998688697815},{"id":"https://openalex.org/keywords/class","display_name":"Class (philosophy)","score":0.47699999809265137},{"id":"https://openalex.org/keywords/alternating-decision-tree","display_name":"Alternating decision tree","score":0.4399000108242035}],"concepts":[{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.8217999935150146},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.765999972820282},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.7425000071525574},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6672000288963318},{"id":"https://openalex.org/C46686674","wikidata":"https://www.wikidata.org/wiki/Q466303","display_name":"Boosting (machine learning)","level":2,"score":0.59579998254776},{"id":"https://openalex.org/C168065819","wikidata":"https://www.wikidata.org/wiki/Q845566","display_name":"Debugging","level":2,"score":0.579200029373169},{"id":"https://openalex.org/C175309249","wikidata":"https://www.wikidata.org/wiki/Q725864","display_name":"Pipeline transport","level":2,"score":0.5561000108718872},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5275999903678894},{"id":"https://openalex.org/C60908668","wikidata":"https://www.wikidata.org/wiki/Q690207","display_name":"Perceptron","level":3,"score":0.508899986743927},{"id":"https://openalex.org/C70153297","wikidata":"https://www.wikidata.org/wiki/Q5591907","display_name":"Gradient boosting","level":3,"score":0.48249998688697815},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.47699999809265137},{"id":"https://openalex.org/C120136583","wikidata":"https://www.wikidata.org/wiki/Q4736414","display_name":"Alternating decision tree","level":5,"score":0.4399000108242035},{"id":"https://openalex.org/C130963320","wikidata":"https://www.wikidata.org/wiki/Q1401207","display_name":"Root cause analysis","level":2,"score":0.4041999876499176},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.40130001306533813},{"id":"https://openalex.org/C84525736","wikidata":"https://www.wikidata.org/wiki/Q831366","display_name":"Decision tree","level":2,"score":0.40059998631477356},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3946000039577484},{"id":"https://openalex.org/C179717631","wikidata":"https://www.wikidata.org/wiki/Q2991667","display_name":"Multilayer perceptron","level":3,"score":0.39320001006126404},{"id":"https://openalex.org/C2777615720","wikidata":"https://www.wikidata.org/wiki/Q11888847","display_name":"Prioritization","level":2,"score":0.39079999923706055},{"id":"https://openalex.org/C84945661","wikidata":"https://www.wikidata.org/wiki/Q7366567","display_name":"Root cause","level":2,"score":0.36500000953674316},{"id":"https://openalex.org/C45804977","wikidata":"https://www.wikidata.org/wiki/Q7239673","display_name":"Predictive modelling","level":2,"score":0.32010000944137573},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.3131999969482422},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.30970001220703125},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.3091999888420105},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.29829999804496765},{"id":"https://openalex.org/C10485038","wikidata":"https://www.wikidata.org/wiki/Q48996162","display_name":"Hyperparameter optimization","level":3,"score":0.2838999927043915},{"id":"https://openalex.org/C45942800","wikidata":"https://www.wikidata.org/wiki/Q245652","display_name":"Ensemble learning","level":2,"score":0.28290000557899475}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1007/s11334-026-00642-8","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11334-026-00642-8","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11334-026-00642-8.pdf","source":{"id":"https://openalex.org/S151239915","display_name":"Innovations in Systems and Software Engineering","issn_l":"1614-5046","issn":["1614-5046","1614-5054"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Innovations in Systems and Software Engineering","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1007/s11334-026-00642-8","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11334-026-00642-8","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11334-026-00642-8.pdf","source":{"id":"https://openalex.org/S151239915","display_name":"Innovations in Systems and Software Engineering","issn_l":"1614-5046","issn":["1614-5046","1614-5054"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Innovations in Systems and Software Engineering","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321870","display_name":"Georg-August-Universit\u00e4t G\u00f6ttingen","ror":"https://ror.org/01y9bpm73"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7165131345.pdf","grobid_xml":"https://content.openalex.org/works/W7165131345.grobid-xml"},"referenced_works_count":19,"referenced_works":["https://openalex.org/W102369970","https://openalex.org/W1575103269","https://openalex.org/W2003984511","https://openalex.org/W2102539288","https://openalex.org/W2103810072","https://openalex.org/W2118978333","https://openalex.org/W2125877832","https://openalex.org/W2132862423","https://openalex.org/W2167467747","https://openalex.org/W2282821441","https://openalex.org/W2739892638","https://openalex.org/W2766412091","https://openalex.org/W2964054038","https://openalex.org/W2973136425","https://openalex.org/W2999752750","https://openalex.org/W3204805998","https://openalex.org/W4317748910","https://openalex.org/W4402937549","https://openalex.org/W4417250719"],"related_works":[],"abstract_inverted_index":{"Abstract":[0],"Machine":[1,123],"learning":[2,395],"pipeline":[3,106,285],"debugging":[4,393],"remains":[5],"costly":[6],"because":[7],"predictive":[8,130],"performance":[9,31,149,352],"depends":[10],"on":[11,62,86],"complex":[12],"interactions":[13],"between":[14],"dataset":[15,136],"characteristics":[16],"and":[17,33,56,76,93,112,128,193,234,246,287,371,387],"configuration":[18,317],"choices.":[19],"In":[20,289],"practice,":[21],"however,":[22],"practitioners":[23],"often":[24,313],"lack":[25],"tools":[26],"that":[27,59,269,378],"can":[28],"both":[29,284,369],"explain":[30],"failures":[32,303],"estimate":[34],"the":[35,84,154,262,334,344,379],"likely":[36],"impact":[37],"of":[38,283,365],"repairs":[39],"without":[40],"repeated":[41],"reruns.":[42],"To":[43],"address":[44],"this":[45],"challenge,":[46],"we":[47,100],"propose":[48],"a":[49,102,126,360],"model-driven":[50],"framework":[51,66,381],"for":[52,232,259,353,368,392],"root":[53,276,357,384],"cause":[54,277,385],"analysis":[55],"hyperparameter":[57,114,346,389],"intervention":[58,347,363,390],"operates":[60],"solely":[61],"structured":[63],"descriptors.":[64],"The":[65,141],"uses":[67],"three":[68,144],"dataset-complexity":[69],"meta-features,":[70],"namely":[71],"class":[72,74,300,309],"overlap,":[73,301,312],"imbalance,":[75],"sparsity,":[77],"together":[78],"with":[79,271,298,333,359],"learner":[80],"hyperparameters.":[81,327],"We":[82,116],"evaluate":[83],"approach":[85],"two":[87,155],"model":[88,156],"families,":[89],"Decision":[90],"Trees":[91],"(DT)":[92],"Multilayer":[94],"Perceptrons":[95],"(MLP).":[96],"For":[97],"each":[98],"family,":[99],"construct":[101],"meta-dataset":[103],"comprising":[104],"81,000":[105],"runs":[107],"generated":[108],"from":[109],"270":[110],"datasets":[111],"300":[113],"configurations.":[115],"then":[117],"train":[118],"an":[119],"interpretable":[120],"Explainable":[121],"Boosting":[122],"(EBM)":[124],"as":[125,320,356],"meta-model":[127,264],"assess":[129],"fidelity":[131],"using":[132],"GroupKFold":[133],"partitioning":[134],"by":[135,315],"to":[137,279,307],"prevent":[138],"data":[139],"leakage.":[140],"study":[142],"yields":[143],"main":[145],"findings.":[146],"First,":[147],"descriptor-based":[148],"prediction":[150,339],"generalizes":[151],"well":[152],"across":[153],"families":[157],"considered,":[158],"achieving":[159],"$$Mean~Absolute~Error":[160],"(MAE)=0.080$$":[161],"<mml:math":[162,196,236,248],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\">":[163,197,237,249],"<mml:mrow>":[164,198,238,250],"<mml:mi>M</mml:mi>":[165,185,204,224,239,252],"<mml:mi>e</mml:mi>":[166,177,205,214],"<mml:mi>a</mml:mi>":[167,206,212],"<mml:mi>n</mml:mi>":[168,207],"<mml:mspace/>":[169,178,203,208,216],"<mml:mi>A</mml:mi>":[170,186,240],"<mml:mi>b</mml:mi>":[171],"<mml:mi>s</mml:mi>":[172],"<mml:mi>o</mml:mi>":[173,182,200,201,220],"<mml:mi>l</mml:mi>":[174],"<mml:mi>u</mml:mi>":[175,211],"<mml:mi>t</mml:mi>":[176,202],"<mml:mi>E</mml:mi>":[179,187,217,226,241,254],"<mml:mi>r</mml:mi>":[180,181,183,213,218,219,221],"<mml:mo>(</mml:mo>":[184,222],"<mml:mo>)</mml:mo>":[188,227],"<mml:mo>=</mml:mo>":[189,228,242,255],"<mml:mn>0.080</mml:mn>":[190],"</mml:mrow>":[191,230,244,257],"</mml:math>":[192,231,245,258],"$$Root~Mean~Squared~Error":[194],"(RMSE)=0.108$$":[195],"<mml:mi>R</mml:mi>":[199,223,251],"<mml:mi>S</mml:mi>":[209,225,253],"<mml:mi>q</mml:mi>":[210],"<mml:mi>d</mml:mi>":[215],"<mml:mn>0.108</mml:mn>":[229],"DT,":[233],"$$MAE=0.087$$":[235],"<mml:mn>0.087</mml:mn>":[243],"$$RMSE=0.115$$":[247],"<mml:mn>0.115</mml:mn>":[256],"MLP.":[260,372],"Second,":[261],"trained":[263],"learns":[265],"directionally":[266],"coherent":[267],"relationships":[268],"align":[270],"prior":[272],"domain":[273],"knowledge,":[274],"enabling":[275],"diagnosis":[278,386],"identify":[280],"meaningful":[281],"drivers":[282],"success":[286],"failure.":[288],"our":[290],"results,":[291],"successful":[292],"ML":[293],"pipelines":[294],"are":[295,304,341],"primarily":[296],"associated":[297],"low":[299],"whereas":[302],"typically":[305],"linked":[306],"high":[308],"imbalance":[310],"or":[311,324],"amplified":[314],"unfavorable":[316],"choices":[318],"such":[319],"restrictive":[321],"tree-growth":[322],"settings":[323],"optimization-sensitive":[325],"MLP":[326],"Importantly,":[328],"these":[329,375],"attributions":[330],"remain":[331],"consistent":[332],"expectation":[335],"rules":[336],"even":[337],"when":[338],"residuals":[340],"non-trivial.":[342],"Third,":[343],"causal":[345],"module":[348],"accurately":[349],"estimates":[350],"post-intervention":[351],"hyperparameters":[354],"identified":[355],"causes,":[358],"mean":[361],"absolute":[362],"error":[364],"approximately":[366],"0.036":[367],"DT":[370],"Taken":[373],"together,":[374],"results":[376],"show":[377],"proposed":[380],"provides":[382],"actionable":[383],"reliable":[388],"guidance":[391],"machine":[394],"pipelines.":[396]},"counts_by_year":[],"updated_date":"2026-06-20T20:08:15.867695","created_date":"2026-06-19T00:00:00"}