{"id":"https://openalex.org/W6943150051","doi":"https://doi.org/10.15167/carlevaro-alberto_phd2024-03-11","title":"Mathematical Methods for eXplainable and Reliable Machine Learning in Trustworthy Artificial Intelligence","display_name":"Mathematical Methods for eXplainable and Reliable Machine Learning in Trustworthy Artificial Intelligence","publication_year":2024,"publication_date":"2024-03-11","ids":{"openalex":"https://openalex.org/W6943150051","doi":"https://doi.org/10.15167/carlevaro-alberto_phd2024-03-11"},"language":"en","primary_location":{"id":"pmh:oai:iris.unige.it:11567/1164117","is_oa":true,"landing_page_url":"https://hdl.handle.net/11567/1164117","pdf_url":null,"source":{"id":"https://openalex.org/S4377196291","display_name":"CINECA IRIS Institutial Research Information System (University of Genoa)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I83816512","host_organization_name":"University of Genoa","host_organization_lineage":["https://openalex.org/I83816512"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"info:eu-repo/semantics/doctoralThesis"},"type":"other","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://hdl.handle.net/11567/1164117","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"CARLEVARO, ALBERTO","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"CARLEVARO, ALBERTO","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T10451","display_name":"Mycorrhizal Fungi and Plant Interactions","score":0.48750001192092896,"subfield":{"id":"https://openalex.org/subfields/1110","display_name":"Plant Science"},"field":{"id":"https://openalex.org/fields/11","display_name":"Agricultural and Biological Sciences"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},"topics":[{"id":"https://openalex.org/T10451","display_name":"Mycorrhizal Fungi and Plant Interactions","score":0.48750001192092896,"subfield":{"id":"https://openalex.org/subfields/1110","display_name":"Plant Science"},"field":{"id":"https://openalex.org/fields/11","display_name":"Agricultural and Biological Sciences"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T10015","display_name":"Genomics and Phylogenetic Studies","score":0.08829999715089798,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T10825","display_name":"Plant Pathogens and Fungal Diseases","score":0.06729999929666519,"subfield":{"id":"https://openalex.org/subfields/1307","display_name":"Cell Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/probabilistic-logic","display_name":"Probabilistic logic","score":0.4410000145435333},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.28929999470710754},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.26100000739097595},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.2485000044107437},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.24539999663829803}],"concepts":[{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.4410000145435333},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41749998927116394},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.28929999470710754},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.26100000739097595},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.25859999656677246},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.2485000044107437},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.24539999663829803},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.24079999327659607},{"id":"https://openalex.org/C2908647359","wikidata":"https://www.wikidata.org/wiki/Q2625603","display_name":"Population","level":2,"score":0.24070000648498535},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.2371000051498413}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:iris.unige.it:11567/1164117","is_oa":true,"landing_page_url":"https://hdl.handle.net/11567/1164117","pdf_url":null,"source":{"id":"https://openalex.org/S4377196291","display_name":"CINECA IRIS Institutial Research Information System (University of Genoa)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I83816512","host_organization_name":"University of Genoa","host_organization_lineage":["https://openalex.org/I83816512"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"info:eu-repo/semantics/doctoralThesis"},{"id":"doi:10.15167/carlevaro-alberto_phd2024-03-11","is_oa":true,"landing_page_url":"https://doi.org/10.15167/carlevaro-alberto_phd2024-03-11","pdf_url":null,"source":{"id":"https://openalex.org/S7407050993","display_name":"Universit\u00e0 degli Studi di Genova","issn_l":null,"issn":[],"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"pmh:oai:iris.unige.it:11567/1164117","is_oa":true,"landing_page_url":"https://hdl.handle.net/11567/1164117","pdf_url":null,"source":{"id":"https://openalex.org/S4377196291","display_name":"CINECA IRIS Institutial Research Information System (University of Genoa)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I83816512","host_organization_name":"University of Genoa","host_organization_lineage":["https://openalex.org/I83816512"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"info:eu-repo/semantics/doctoralThesis"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"The":[0,42,148,632,736,765],"common":[1],"thread":[2],"in":[3,17,54,89,173,239,289,331,354,415,446,478,571,617,667,705,709,720,729,778,806,838,850,863,906],"all":[4,284,802,846],"my":[5,75,112,226,440,447,572,734,847,856],"research":[6,76,457,557,848,862],"has":[7,395,809],"been":[8,396,810],"the":[9,18,31,34,40,50,57,85,105,145,156,174,186,189,199,232,240,243,259,266,290,301,305,328,362,371,375,387,390,460,464,486,508,512,533,537,541,544,566,608,618,645,657,661,668,680,687,696,700,710,723,730,750,753,794,803,815,864,872,875,896,902],"goal":[10,58],"of":[11,21,33,44,52,84,150,188,208,228,234,242,245,268,275,321,339,361,377,439,456,466,511,536,546,556,565,596,610,620,634,648,689,712,725,733,752,817,840,861,866,871,877,893,898,904,911],"finding":[12],"a":[13,79,136,139,167,180,204,221,249,272,286,435,470,525,551,611,623,649,691,713,755,881,887,908],"so-called":[14],"safety":[15,45,157,209,229,260,269,306,322,363,414,461,726,800],"region":[16,46,169,230,270,307,462,727],"input":[19,175,646,698],"space":[20,176],"an":[22,121,152,499],"inference":[23],"model":[24,35,513,538],"that":[25,182,283,294,317,492,507,531,548,638,663,679],"allows":[26,166,325,783],"providing":[27],"probabilistic":[28,70,235,256,359,393],"guarantees":[29,257,303,360],"on":[30,258,386,550,600,717,759],"output":[32,510,609],"and":[36,65,81,141,163,177,251,308,356,427,539,562,602,699,824,842,890],"tools":[37],"to":[38,60,96,110,131,170,192,198,220,254,271,299,326,336,351,357,400,409,412,433,496,501,523,543,579,591,606,642,655,671,686,742,770,820,879],"control":[39,185,352,366,607,777],"prediction.":[41],"idea":[43,227,267,487,633,688,724],"fits":[47],"well":[48],"with":[49,138,211,722,793,852,895],"task":[51],"classification":[53,87,190,281,355,612],"machine":[55,90,479,650,883],"learning:":[56],"is":[59,130,160,491,494,514,521,558,563,576,587,637,640],"classify":[61],"instances":[62],"into":[63,314],"well-defined":[64],"closed":[66,168],"envelopes,":[67],"respecting":[68],"some":[69],"performance":[71,200,216,353],"or":[72,194,380,383],"guarantees.":[73],"So":[74],"started":[77],"from":[78,92,144,203,334,886],"thorough":[80],"accurate":[82],"review":[83],"main":[86,128],"algorithms":[88,349],"learning,":[91],"support":[93],"vector":[94],"machines":[95],"neural":[97],"networks":[98],"via":[99],"rule-based":[100],"models":[101,282],"as":[102,420,515,654,775],"well.":[103],"But":[104],"best":[106],"algorithm":[107,123,153,481,500,542,652,885],"I":[108,218,264,309,346,568,598],"found":[109],"achieve":[111,413],"purpose":[113,129],"was":[114,367,463,615,739,763,768],"Support":[115],"Vector":[116],"Data":[117],"Descrip-":[118],"tion":[119],"(SVDD),":[120],"established":[122],"for":[124,154,304,458,475,784,787],"outlier":[125],"detection":[126],"whose":[127],"enclose":[132],"target":[133],"data":[134,146],"within":[135,231],"sphere":[137],"center":[140],"radius":[142,181],"learned":[143],"distribution.":[147],"choice":[149],"such":[151,419,774,907],"defining":[155,459],"re-":[158],"gion":[159],"quite":[161],"trivial":[162],"supportable:":[164],"SVDD":[165,335],"be":[171,296,401,407,665,677,743,836],"defined":[172],"also":[178,310,835],"provides":[179,248],"can":[183,295,406,834],"easily":[184],"shape":[187],"boundary":[191],"\u201cinflate\u201d":[193],"\u201cdeflate\u201d":[195],"it":[196,493,520,639,791],"according":[197],"objective.":[201],"Starting":[202],"totally":[205,582,719,849],"data-driven":[206],"definition":[207,293],"region,":[210],"only":[212],"empirical":[213],"(but":[214],"effective)":[215],"guarantees,":[217],"moved":[219],"more":[222,785],"mathematical":[223],"definition,":[224],"placing":[225,845],"framework":[233],"scaling.":[236],"This":[237,324,554,614,780],"technique,":[238],"state":[241],"art":[244],"order":[246],"statistics,":[247],"clear":[250,678],"indisputable":[252],"way":[253],"obtain":[255,300,358,502],"region.":[261,364],"Here,":[262],"moreover,":[263],"applied":[265,408,769],"broader":[273],"class":[274],"classifiers,":[276,279],"called":[277,528],"scalable":[278,287],"i.e.,":[280],"share":[285],"parameter":[288],"classifier\u2019s":[291],"predictor":[292],"appropriately":[297],"adjusted":[298],"desired":[302,509,701],"specialized":[311],"these":[312],"concepts":[313,329],"exponential":[315],"distributions":[316],"allow":[318],"special":[319],"properties":[320],"regions.":[323],"extend":[327],"developed":[330],"Chapter":[332],"3":[333],"any":[337],"kind":[338],"ma-":[340],"chine":[341],"learning":[342,480,651,884],"classifier.":[343],"In":[344,483,517,584,660,829],"particular,":[345],"introduced":[347],"new":[348,471,604,625],"both":[350],"Performance":[365],"achieved":[368],"by":[369,789,799,812,826],"minimizing":[370,690],"misclassification":[372],"error,":[373],"reducing":[374],"number":[376],"false":[378,381],"positives":[379],"negatives":[382],"both,":[384],"depending":[385],"application.":[388],"On":[389],"other":[391,454],"hand,":[392],"guarantee":[394],"shown":[397],"mathemat-":[398],"ically":[399],"effective.":[402],"Both":[403],"concepts,":[404],"however,":[405,782],"real-world":[410,771],"problems":[411],"cyber-physical":[416],"systems":[417],"applications,":[418],"vehicle":[421],"platooning":[422],"monitoring,":[423],"DNS":[424],"tunneling":[425],"detection,":[426],"type-2":[428],"diabetes":[429],"disease":[430],"prediction,":[431],"just":[432],"name":[434],"few":[436],"tested":[437],"applications":[438],"methods.
\\nHowever,":[441],"before":[442],"getting":[443,559],"good":[444,560,882],"results":[445],"research,":[448],"several":[449],"ways":[450],"were":[451],"tried.":[452],"An-":[453],"line":[455,555,721,851],"use":[465],"conformal":[467,489,795],"predic-":[468],"tion,":[469],"but":[472,626,747],"well-established":[473],"theory":[474],"evaluating":[476,547,874],"conformity":[477],"performance.":[482],"this":[484,518,594,706,807,830,853],"case,":[485],"behind":[488],"prediction":[490,658],"possible":[495,641],"correctly":[497],"calibrate":[498,540],"marginal":[503],"probability":[504],"cover-":[505],"age":[506],"expected.":[516],"field,":[519],"necessary":[522],"define":[524],"real-valued":[526],"function,":[527,530],"score":[529],"encodes":[532],"characteris-":[534],"tics":[535],"result":[545],"function":[549,694],"calibration":[552],"set.":[553],"prospects":[561],"one":[564],"lines":[567],"will":[569,664,676],"follow":[570],"future":[573],"work.
\\nBut":[574],"reliability":[575],"not":[577],"enough":[578],"make":[580,880],"AI":[581,630,822,833],"trustworthy.":[583],"fact,":[585],"controllability":[586,841],"another":[588],"crucial":[589,909],"aspect":[590,910],"consider.":[592],"From":[593],"point":[595,892],"view,":[597,894],"focused":[599],"studying":[601],"developing":[603],"techniques":[605,762],"algorithm.":[613],"done":[616],"spirit":[619],"counterfactual":[621,635,672,714],"explanation,":[622],"fairly":[624],"already":[627],"state-of-the-art":[628],"eXplanaible":[629],"technique.":[631],"explanations":[636,673],"minimally":[643],"change":[644,656],"parameters":[647],"so":[653],"results.":[659],"sense":[662],"explained":[666],"chapter":[669],"dedicated":[670,819],"(Chapter":[674],"9)":[675],"expression":[681],"\u201c":[682],"minimal":[683],"change\u201d":[684],"refers":[685],"specific":[692],"cost":[693],"between":[695],"actual":[697],"one.":[702],"My":[703],"contribution":[704],"topic":[707],"lies":[708],"development":[711],"approach":[715,738],"based":[716,758],"SVDD,":[718],"investigated":[728],"first":[731,740],"part":[732],"research.":[735],"proposed":[737],"attempted":[741],"solved":[744],"completely":[745],"analytically,":[746],"then,":[748],"given":[749],"complexity":[751],"task,":[754],"numerical":[756],"solution":[757],"random":[760],"sampling":[761],"developed.":[764],"algorithm,":[766],"again,":[767],"application":[772],"problems,":[773],"crowd":[776],"subways.":[779],"topic,":[781],"exploration,":[786],"example":[788],"merging":[790],"together":[792],"frame-":[796],"work":[797,804],"provided":[798],"regions.
\\nFinally,":[801],"presented":[805],"thesis":[808,857],"surrounded":[811],"explainable":[813,823,832,889],"AI,":[814],"field":[816,865],"study":[818],"making":[821],"expressible":[825],"intelligible":[827],"rules.":[828],"regard,":[831],"declined":[837],"terms":[839],"reliability,":[843],"thus":[844],"theme.
\\nIn":[854],"conclusion,":[855],"covered":[858],"three":[859],"years":[860],"artificial":[867],"intelligence,":[868],"spending":[869],"most":[870],"time":[873],"problem":[876],"how":[878],"reliable,":[888],"controllable":[891],"hope":[897],"having":[899],"really":[900],"improved":[901],"body":[903],"knowledge":[905],"Science.":[912]},"counts_by_year":[],"updated_date":"2026-02-09T09:26:11.010843","created_date":"2025-10-10T00:00:00"}