{"id":"https://openalex.org/W4363672014","doi":"https://doi.org/10.48550/arxiv.2304.03486","title":"Can we learn better with hard samples?","display_name":"Can we learn better with hard samples?","alternate_titles":[],"publication_year":2023,"publication_date":"2023-04-07","ids":{"openalex":"https://openalex.org/W4363672014","doi":"https://doi.org/10.48550/arxiv.2304.03486"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2304.03486","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2304.03486","pdf_url":"https://arxiv.org/pdf/2304.03486","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"parent_works":[],"lineage":["https://openalex.org/W4363672014"],"child_works_count":0,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2304.03486","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5082738236","display_name":"Subin Sahayam","orcid":"https://orcid.org/0000-0003-1129-895X","observed_orcids":["https://orcid.org/0000-0003-1129-895X"]},"institutions":[{"id":"https://openalex.org/I39244652","display_name":"Indian Institute of Information Technology, Design and Manufacturing, Kancheepuram","ror":"https://ror.org/023c9pb11","country_code":"IN","type":"education","lineage":["https://openalex.org/I39244652"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Sahayam, Subin","raw_affiliation_strings":["Indian Institute of Information Technolgy , Kancheepuram"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Indian Institute of Information Technolgy , Kancheepuram","institution_ids":["https://openalex.org/I39244652"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051260256","display_name":"John Zakkam","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I39244652","display_name":"Indian Institute of Information Technology, Design and Manufacturing, Kancheepuram","ror":"https://ror.org/023c9pb11","country_code":"IN","type":"education","lineage":["https://openalex.org/I39244652"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Zakkam, John","raw_affiliation_strings":["Indian Institute of Information Technolgy , Kancheepuram"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Indian Institute of Information Technolgy , Kancheepuram","institution_ids":["https://openalex.org/I39244652"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5058647078","display_name":"Umarani Jayaraman","orcid":"https://orcid.org/0000-0002-9676-6291","observed_orcids":["https://orcid.org/0000-0002-9676-6291"]},"institutions":[{"id":"https://openalex.org/I39244652","display_name":"Indian Institute of Information Technology, Design and Manufacturing, Kancheepuram","ror":"https://ror.org/023c9pb11","country_code":"IN","type":"education","lineage":["https://openalex.org/I39244652"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Jayaraman, Umarani","raw_affiliation_strings":["Indian Institute of Information Technolgy , Kancheepuram"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Indian Institute of Information Technolgy , Kancheepuram","institution_ids":["https://openalex.org/I39244652"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I39244652"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":5,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_pulled":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.4325999915599823,"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":0.4325999915599823,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.33640000224113464,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.13179999589920044,"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/hard-sample-mining","display_name":"hard sample mining","score":0.6679999828338623},{"id":"https://openalex.org/keywords/mini-batch-training","display_name":"mini-batch training","score":0.8809999823570251},{"id":"https://openalex.org/keywords/class-imbalance","display_name":"class imbalance","score":0.5130000114440918},{"id":"https://openalex.org/keywords/generalization","display_name":"generalization","score":0.48899999260902405},{"id":"https://openalex.org/keywords/convergence-rate","display_name":"convergence rate","score":0.7089999914169312},{"id":"https://openalex.org/keywords/image-classification","display_name":"image classification","score":0.5170000195503235},{"id":"https://openalex.org/keywords/cifar-10","display_name":"CIFAR-10","score":0.8640000224113464},{"id":"https://openalex.org/keywords/cifar-100","display_name":"CIFAR-100","score":0.9980000257492065},{"id":"https://openalex.org/keywords/stl-10","display_name":"STL-10","score":0.925000011920929},{"id":"https://openalex.org/keywords/resnet","display_name":"ResNet","score":0.8539999723434448},{"id":"https://openalex.org/keywords/efficientnet","display_name":"EfficientNet","score":0.8220000267028809},{"id":"https://openalex.org/keywords/mobilenetv3","display_name":"MobileNetV3","score":0.9850000143051147}],"concepts":[{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7497530579566956},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.7256150245666504},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7189092040061951},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6576427817344666},{"id":"https://openalex.org/C2944601119","wikidata":"https://www.wikidata.org/wiki/Q43744058","display_name":"Residual neural network","level":3,"score":0.6507304310798645},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5436363816261292},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.5241898894309998},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.50596684217453},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.48451298475265503},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.44977307319641113},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.44426149129867554},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41790932416915894},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.39432623982429504},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3517571985721588},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.21101167798042297},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C50522688","wikidata":"https://www.wikidata.org/wiki/Q189833","display_name":"Economic growth","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2304.03486","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2304.03486","pdf_url":"https://arxiv.org/pdf/2304.03486","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2304.03486","doi":"https://doi.org/10.48550/arxiv.2304.03486","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2304.03486","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2304.03486","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2304.03486","pdf_url":"https://arxiv.org/pdf/2304.03486","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4363672014.pdf","grobid_xml":"https://content.openalex.org/works/W4363672014.grobid-xml"},"referenced_works_count":48,"referenced_works":["https://openalex.org/W114517082","https://openalex.org/W1498436455","https://openalex.org/W1573503290","https://openalex.org/W1686810756","https://openalex.org/W1766594731","https://openalex.org/W2020999234","https://openalex.org/W2040870580","https://openalex.org/W2053717624","https://openalex.org/W2100495367","https://openalex.org/W2106220446","https://openalex.org/W2118858186","https://openalex.org/W2163605009","https://openalex.org/W2163922914","https://openalex.org/W2194775991","https://openalex.org/W2338318698","https://openalex.org/W2473611539","https://openalex.org/W2518108298","https://openalex.org/W2523246573","https://openalex.org/W2682189153","https://openalex.org/W2809598685","https://openalex.org/W2897025488","https://openalex.org/W2912323147","https://openalex.org/W2919115771","https://openalex.org/W2925140953","https://openalex.org/W2936503027","https://openalex.org/W2944779197","https://openalex.org/W2945020349","https://openalex.org/W2945785363","https://openalex.org/W2954996726","https://openalex.org/W2963351448","https://openalex.org/W2963446712","https://openalex.org/W2963460174","https://openalex.org/W2964121744","https://openalex.org/W2965857035","https://openalex.org/W2970971581","https://openalex.org/W2979417040","https://openalex.org/W2998768810","https://openalex.org/W3011655242","https://openalex.org/W3045004532","https://openalex.org/W3094502228","https://openalex.org/W3098033339","https://openalex.org/W3138516171","https://openalex.org/W3145444543","https://openalex.org/W3182797138","https://openalex.org/W4231109964","https://openalex.org/W4288083766","https://openalex.org/W4324142429","https://openalex.org/W7225208346"],"related_works":["https://openalex.org/W2799042347","https://openalex.org/W7030195791","https://openalex.org/W3001471657","https://openalex.org/W4212801704","https://openalex.org/W2883470566","https://openalex.org/W2588722418","https://openalex.org/W4414396971","https://openalex.org/W2773689216","https://openalex.org/W7133195748","https://openalex.org/W2997599645"],"abstract_inverted_index":{"In":[0],"deep":[1,70,83],"learning,":[2],"mini-batch":[3,15,122,184],"training":[4,67,123],"is":[5],"commonly":[6],"used":[7,86],"to":[8,30,119],"optimize":[9],"network":[10,52],"parameters.":[11],"However,":[12],"the":[13,20,27,43,51,62,65,88,104,110,116,120,149,152,171,182,195],"traditional":[14,44,121,183],"method":[16,106,154,173,185,192],"may":[17],"not":[18],"learn":[19],"under-represented":[21],"samples":[22],"and":[23,80,97,113,165],"complex":[24],"patterns":[25],"in":[26,87,161,175,186,201],"data,":[28],"leading":[29],"a":[31,40,128],"longer":[32],"time":[33],"for":[34,139,155],"generalization.":[35,167],"To":[36],"address":[37],"this":[38],"problem,":[39],"variant":[41],"of":[42,64,145,151,179],"algorithm":[45],"has":[46],"been":[47],"proposed,":[48],"which":[49],"trains":[50],"focusing":[53],"on":[54,74,142,188,203],"mini-batches":[55,136],"with":[56],"high":[57],"loss.":[58],"The":[59,82,99,190],"study":[60,89],"evaluates":[61],"effectiveness":[63],"proposed":[66,105,153,172,191],"using":[68],"various":[69,143],"neural":[71,84],"networks":[72,85],"trained":[73],"three":[75],"benchmark":[76],"datasets":[77],"(CIFAR-10,":[78],"CIFAR-100,":[79],"STL-10).":[81],"are":[90,137],"ResNet-18,":[91],"ResNet-50,":[92],"Efficient":[93],"Net":[94],"B4,":[95],"EfficientNetV2-S,":[96],"MobilenetV3-S.":[98],"experimental":[100],"results":[101,160],"showed":[102],"that":[103,132,148,170],"can":[107],"significantly":[108],"improve":[109],"test":[111,163,196],"accuracy":[112,164,198],"speed":[114],"up":[115],"convergence":[117],"compared":[118],"method.":[124],"Furthermore,":[125],"we":[126],"introduce":[127],"hyper-parameter":[129],"delta":[130],"(\u03b4)":[131],"decides":[133],"how":[134],"many":[135],"considered":[138],"training.":[140],"Experiments":[141],"values":[144,158],"\u03b4":[146,157],"found":[147],"performance":[150],"smaller":[156],"generally":[159],"similar":[162],"faster":[166],"We":[168],"show":[169],"generalizes":[174],"26.47%":[176],"less":[177],"number":[178],"epochs":[180],"than":[181],"EfficientNet-B4":[187],"STL-10.":[189],"also":[193],"improves":[194],"top-1":[197],"by":[199],"7.26%":[200],"ResNet-18":[202],"CIFAR-100.":[204]},"cited_by_api_url":"https://api.openalex.org/works?filter=cites:W4363672014","counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":2}],"updated_date":"2026-11-16T12:05:36.265796","created_date":"2023-04-11T00:00:00"}
