{"id":"https://openalex.org/W7164822767","doi":"https://doi.org/10.3390/a19060482","title":"Optimize Multimodal Data Mixture for Pre-Training with Loss Regression","display_name":"Optimize Multimodal Data Mixture for Pre-Training with Loss Regression","publication_year":2026,"publication_date":"2026-06-15","ids":{"openalex":"https://openalex.org/W7164822767","doi":"https://doi.org/10.3390/a19060482"},"language":"en","primary_location":{"id":"doi:10.3390/a19060482","is_oa":true,"landing_page_url":"https://doi.org/10.3390/a19060482","pdf_url":"https://www.mdpi.com/1999-4893/19/6/482/pdf?version=1781530766","source":{"id":"https://openalex.org/S190629608","display_name":"Algorithms","issn_l":"1999-4893","issn":["1999-4893"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310310987","host_organization_name":"Multidisciplinary Digital Publishing Institute","host_organization_lineage":["https://openalex.org/P4310310987"],"host_organization_lineage_names":["Multidisciplinary Digital Publishing Institute"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Algorithms","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.mdpi.com/1999-4893/19/6/482/pdf?version=1781530766","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5121384952","display_name":"Linjiang Shang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210126065","display_name":"Shanghai Electric (China)","ror":"https://ror.org/0314qy595","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210126065"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Linjiang Shang","raw_affiliation_strings":["Information and Communication Branch, State Grid Jiangsu Electric Power Company Ltd., Nanjing 210024, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Information and Communication Branch, State Grid Jiangsu Electric Power Company Ltd., Nanjing 210024, China","institution_ids":["https://openalex.org/I4210126065"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068753535","display_name":"Huanyu Cheng","orcid":"https://orcid.org/0000-0003-3421-9556"},"institutions":[{"id":"https://openalex.org/I4210126065","display_name":"Shanghai Electric (China)","ror":"https://ror.org/0314qy595","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210126065"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huanyu Cheng","raw_affiliation_strings":["Information and Communication Branch, State Grid Jiangsu Electric Power Company Ltd., Nanjing 210024, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Information and Communication Branch, State Grid Jiangsu Electric Power Company Ltd., Nanjing 210024, China","institution_ids":["https://openalex.org/I4210126065"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100335410","display_name":"Bo Zhou","orcid":"https://orcid.org/0009-0000-5724-0951"},"institutions":[{"id":"https://openalex.org/I168879160","display_name":"Zhejiang University of Science and Technology","ror":"https://ror.org/05mx0wr29","country_code":"CN","type":"education","lineage":["https://openalex.org/I168879160"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Bo Zhou","raw_affiliation_strings":["College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China"],"raw_orcid":"https://orcid.org/0009-0000-5724-0951","affiliations":[{"raw_affiliation_string":"College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China","institution_ids":["https://openalex.org/I168879160"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5068103926","display_name":"Y Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I168879160","display_name":"Zhejiang University of Science and Technology","ror":"https://ror.org/05mx0wr29","country_code":"CN","type":"education","lineage":["https://openalex.org/I168879160"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yin Zhang","raw_affiliation_strings":["College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China","institution_ids":["https://openalex.org/I168879160"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":{"value":1600,"currency":"CHF","value_usd":1782},"apc_paid":{"value":1600,"currency":"CHF","value_usd":1782},"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.75484292,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"19","issue":"6","first_page":"482","last_page":"482"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13702","display_name":"Machine Learning in Healthcare","score":0.21359999477863312,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.21359999477863312,"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/T10028","display_name":"Topic Modeling","score":0.15549999475479126,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.09279999881982803,"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/smoothing","display_name":"Smoothing","score":0.6916999816894531},{"id":"https://openalex.org/keywords/hyperparameter","display_name":"Hyperparameter","score":0.64410001039505},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.6341999769210815},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.43299999833106995},{"id":"https://openalex.org/keywords/mixture-model","display_name":"Mixture model","score":0.42179998755455017},{"id":"https://openalex.org/keywords/regression","display_name":"Regression","score":0.39259999990463257},{"id":"https://openalex.org/keywords/synthetic-data","display_name":"Synthetic data","score":0.3912999927997589},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.3357999920845032}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7907999753952026},{"id":"https://openalex.org/C3770464","wikidata":"https://www.wikidata.org/wiki/Q775963","display_name":"Smoothing","level":2,"score":0.6916999816894531},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.64410001039505},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.6341999769210815},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5418999791145325},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5027999877929688},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.43299999833106995},{"id":"https://openalex.org/C61224824","wikidata":"https://www.wikidata.org/wiki/Q2260434","display_name":"Mixture model","level":2,"score":0.42179998755455017},{"id":"https://openalex.org/C83546350","wikidata":"https://www.wikidata.org/wiki/Q1139051","display_name":"Regression","level":2,"score":0.39259999990463257},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.3912999927997589},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3709999918937683},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3357999920845032},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.33169999718666077},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.3091999888420105},{"id":"https://openalex.org/C21080849","wikidata":"https://www.wikidata.org/wiki/Q13611879","display_name":"Data point","level":2,"score":0.30410000681877136},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.28380000591278076},{"id":"https://openalex.org/C2780660688","wikidata":"https://www.wikidata.org/wiki/Q25052564","display_name":"Multimodal learning","level":2,"score":0.27649998664855957},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2678999900817871},{"id":"https://openalex.org/C56672385","wikidata":"https://www.wikidata.org/wiki/Q17157111","display_name":"Mixture distribution","level":3,"score":0.2551000118255615},{"id":"https://openalex.org/C48921125","wikidata":"https://www.wikidata.org/wiki/Q10861030","display_name":"Linear regression","level":2,"score":0.2533000111579895}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.3390/a19060482","is_oa":true,"landing_page_url":"https://doi.org/10.3390/a19060482","pdf_url":"https://www.mdpi.com/1999-4893/19/6/482/pdf?version=1781530766","source":{"id":"https://openalex.org/S190629608","display_name":"Algorithms","issn_l":"1999-4893","issn":["1999-4893"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310310987","host_organization_name":"Multidisciplinary Digital Publishing Institute","host_organization_lineage":["https://openalex.org/P4310310987"],"host_organization_lineage_names":["Multidisciplinary Digital Publishing Institute"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Algorithms","raw_type":"journal-article"},{"id":"pmh:oai:doaj.org/article:7d84ed51e2854988921adcdad412c4b8","is_oa":false,"landing_page_url":"https://doaj.org/article/7d84ed51e2854988921adcdad412c4b8","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Algorithms, Vol 19, Iss 6, p 482 (2026)","raw_type":"article"}],"best_oa_location":{"id":"doi:10.3390/a19060482","is_oa":true,"landing_page_url":"https://doi.org/10.3390/a19060482","pdf_url":"https://www.mdpi.com/1999-4893/19/6/482/pdf?version=1781530766","source":{"id":"https://openalex.org/S190629608","display_name":"Algorithms","issn_l":"1999-4893","issn":["1999-4893"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310310987","host_organization_name":"Multidisciplinary Digital Publishing Institute","host_organization_lineage":["https://openalex.org/P4310310987"],"host_organization_lineage_names":["Multidisciplinary Digital Publishing Institute"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Algorithms","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320336329","display_name":"State Grid Jiangsu Electric Power","ror":null}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7164822767.pdf","grobid_xml":"https://content.openalex.org/works/W7164822767.grobid-xml"},"referenced_works_count":15,"referenced_works":["https://openalex.org/W2886641317","https://openalex.org/W2984008963","https://openalex.org/W2989322838","https://openalex.org/W3156892778","https://openalex.org/W3176641147","https://openalex.org/W3204588463","https://openalex.org/W4312299632","https://openalex.org/W4313190371","https://openalex.org/W4398234485","https://openalex.org/W4402670537","https://openalex.org/W4402671548","https://openalex.org/W4402671688","https://openalex.org/W4402716477","https://openalex.org/W4402727885","https://openalex.org/W4404575065"],"related_works":[],"abstract_inverted_index":{"Different":[0],"mixtures":[1,19,188],"of":[2,10,28,74,89,96,118,175,189,197],"multimodal":[3,11,62,120,190],"training":[4,76,191,198],"data":[5,18,77,83,106,121,137,199],"significantly":[6],"impact":[7],"the":[8,45,50,72,119,161,194],"performance":[9,131],"large":[12],"language":[13],"models,":[14],"and":[15,24,41,70,113,133,156,171],"manually":[16],"tuning":[17],"is":[20,80,98,165],"inefficient,":[21],"computationally":[22,183],"expensive,":[23],"frequently":[25],"suboptimal":[26],"because":[27],"complex,":[29],"nonlinear":[30],"inter-modal":[31],"interactions.":[32],"How":[33],"to":[34],"determine":[35],"data-mixture":[36,63],"hyperparameters":[37],"in":[38,49],"an":[39],"efficient":[40,184],"principled":[42],"manner":[43],"becomes":[44],"bottleneck":[46],"for":[47,186],"progress":[48],"field.":[51],"This":[52],"study":[53],"establishes":[54],"a":[55,66,172,181],"scalable,":[56],"learnable":[57],"framework,":[58],"DMPredictor,":[59],"that":[60],"treats":[61],"design":[64],"as":[65],"regression-based":[67],"hyperparameter-optimization":[68],"problem":[69],"automates":[71],"selection":[73],"effective":[75],"mixtures.":[78,107,138],"DMPredictor":[79,128],"trained":[81,99],"on":[82,100,146,151,154,158],"mixture":[84,122,142,162],"samples":[85],"derived":[86],"from":[87],"hundreds":[88],"small":[90,169,173],"proxy":[91],"models":[92],"(2M":[93],"parameters),":[94],"each":[95],"which":[97],"1B":[101],"tokens":[102],"sampled":[103],"using":[104],"different":[105],"The":[108,139,177],"framework":[109],"incorporates":[110],"alignment-aware":[111],"smoothing":[112],"quality-reweighting,":[114],"enabling":[115],"diverse":[116,147],"exploration":[117],"space":[123],"while":[124],"avoiding":[125],"distribution":[126],"collapse.":[127],"produces":[129],"accurate":[130],"forecasts":[132],"identifies":[134],"nearly":[135],"optimal":[136,141],"predicted":[140],"surpasses":[143],"human-designed":[144],"baselines":[145],"benchmarks,":[148],"achieving":[149],"+2.7%":[150],"MMMU,":[152],"+6.4%":[153],"TextVQA,":[155],"+195.2":[157],"MME.":[159],"Moreover,":[160],"optimization":[163],"complexity":[164],"largely":[166],"reduced":[167],"by":[168],"proxies":[170],"number":[174],"tokens.":[176],"proposed":[178],"approach":[179],"offers":[180],"robust,":[182],"pathway":[185],"optimizing":[187],"data,":[192],"addressing":[193],"critical":[195],"challenge":[196],"heterogeneity.":[200]},"counts_by_year":[],"updated_date":"2026-07-23T08:03:31.855105","created_date":"2026-06-16T00:00:00"}