Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
CCCの最終発表
Search
キンジョウショウタロウ
March 04, 2022
0
23
CCCの最終発表
2021に参加さしてもらったときの最終発表スライド(TrendOwl)
キンジョウショウタロウ
March 04, 2022
Tweet
Share
More Decks by キンジョウショウタロウ
See All by キンジョウショウタロウ
逆求人スライド
kinjyo
0
11
夏のゼミ課題
kinjyo
0
16
春学期最終プレゼンスライド
kinjyo
0
25
monthly-LT-LINE.pdf
kinjyo
0
26
NSEEMのLTの資料です
kinjyo
0
80
LINE_DC
kinjyo
0
19
easy_easy_re.pdf
kinjyo
0
240
easy_easy3.pdf
kinjyo
0
22
学校登壇.pdf
kinjyo
0
11
Featured
See All Featured
Optimising Largest Contentful Paint
csswizardry
33
3k
Build The Right Thing And Hit Your Dates
maggiecrowley
33
2.4k
jQuery: Nuts, Bolts and Bling
dougneiner
61
7.5k
How To Stay Up To Date on Web Technology
chriscoyier
789
250k
Building Flexible Design Systems
yeseniaperezcruz
327
38k
Performance Is Good for Brains [We Love Speed 2024]
tammyeverts
6
510
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
226
22k
Distributed Sagas: A Protocol for Coordinating Microservices
caitiem20
330
21k
YesSQL, Process and Tooling at Scale
rocio
169
14k
How to Create Impact in a Changing Tech Landscape [PerfNow 2023]
tammyeverts
48
2.2k
Put a Button on it: Removing Barriers to Going Fast.
kastner
59
3.6k
Fight the Zombie Pattern Library - RWD Summit 2016
marcelosomers
232
17k
Transcript
None
.FNCFST Α͏͔Μ ͿΜͿΜ ͍ʔͪΌΜ 5ZMFS Ωϯδϣ ͱΈͬ͘͢ ࣎լ ژ ౦ژ
ౡ aaࢲ͕ͳ͍ͯ͠·͢
None
IUUQTRJJUBDPNBCDTIPUBSPJUFNTGDCBDF
ເΛඳ͖͍ͨ ʮϑΝΫτνΣοΫͷຽओԽʯ Λࢦ͍ͯ͠·͢ʜ
੍࡞ 🤔ϑΣΠΫχϡʔεͱ
໋࣏ʹؔΘΔ͜ͱ ۽ຊ ถબڍ
ϫΫνϯࣗͷҙࢥͰଧͯΔ ͷʹɺଧͭͳͱڧ੍͞Εͯ͠ ·͍ͬͯΔʜ ࣮ࡍͷମݧஊ ͕ใݯΛͭʹཔΓ͗ͯ͢ɺ ภͬͨใͰ͢Α͏ʹͳΓɺ ձ͕Γཱͨͳ͍ ϝϯόʔͷ5͘Μ ༑ୡͷ/͞Μ
૯লͷൃදͰϲ݄ʹҎ্ͷਓϑΣΠΫχϡʔεʹ৮Ε͍ͯΔ Ҿ༻ɿIUUQTXXXTPVNVHPKQNBJO@TPTJLJKPIP@UTVTJOE@TZPIJJIPZVHBJ@IUNM ʢຊݚڀձʢୈճʣࢿྉ̍ʮʰϑΣΠΫχϡʔεʱʹؔ͢ΔΞϯέʔτௐࠪ݁Ռʯʢ/3*ʣʣ
ίϩφʹؔ͢ΔใͷਅِΛஅͰ͖ͳ͔ͬͨਓ͕ଟ͍ ɾίϩφՒͷຊͰɺ ʮ͓౬Ͱ৽ܕίϩφΠϧε͕ࢮ໓͢Δʯ ʮ৽ܕίϩφΠϧεͷӨڹͰ τΠϨοτϖʔύʔ͕ෆ͢Δʯ ͳͲͷِใ͕֦ࢄͨ͠ɻ
ɾ૯লͷௐࠪͰɺ͜ΕΒͷِใͷ͏ͪ ҰͭͰ৴ͨ͡ͱ͑ͨਓ ྸ͕͘ͳΔ΄Ͳଟ͘ɺ ̍̑ʙ̍̕ࡀͰ̏̒ʹٴΜͰ͍Δɻ Ҿ༻ɿIUUQTXXXTPVNVHPKQNBJO@DPOUFOUQEG ʢ৽ܕίϩφΠϧεײછʹؔ͢Δใྲྀ௨ௐࠪʣ
Ҿ༻ɿIUUQTXXXOBUVSFDPNBSUJDMFTT ʢ5IFPOMJOFDPNQFUJUJPOCFUXFFOQSPBOEBOUJWBDDJOBUJPOWJFXTʣ ϫΫνϯର͕͜ͷ··૿͑ଓ͚Δͱ͍ۙকདྷऔΓࠐ·Εͯ͠·͏ةݥੑ͕͋Δͱ͍͏ݚڀ݁Ռ😱
͍ۙকདྷͷةػײ ใੜଶܥͰੜ͞ΕΔใ͕૿େ͠ɺ ͔ͭෆ͔֬ͳใͷׂ߹͕ߴ͍ʢͭ·ΓϊΠζ͕ ଟ͍ʣใաଟͷঢ়ଶ͕ੜ͡Δ͜ͱͰɺِใΛ ࣄ࣮ͱޡೝͨ͠Γɺ͋Δ͍ࣄ࣮Λࣄ࣮ͩͱ৴͡ ΒΕͳ͍Α͏ͳ͜ͱ͕සൟʹى͜ΔΑ͏ʹͳΕ ɺৗੜ׆ܦࡁ׆ಈɺ͞Βʹຽओओٛʹ ෛͷӨڹ͕ٴͿՄೳੑ͕͋Δɻ
ʁ Έͳ͞ΜɺϑΣΠΫχϡʔε ରࡦͬͯ·͔͢ʁ ʁ
͍͍͑ ͍ ΠϯελάϥϜʹͯνʔϜϝϯόʔௐ ʮϑΣΠΫχϡʔεʹԿ͔͠ΒରࡦΛ͍ͯ͠Δ͔ʯ ߴߍੜɾେֶੜਓ
͔͠͠ϑΣΠΫχϡʔε ʹର͢ΔҙࣝΛ͍࣋ͬͯ Δਓ·ͩ·ͩগͳ͍ ௐࠪͷ݁Ռ ϑΣΠΫχϡʔεʹ৮ΕΔසʹʹ૿͍͑ͯΔ
ײओ؍ʹࠨӈ͞Εͣ ٬؍తͳࠜڌ Λ࣋ͬͯΒ͍͍ͨ Ծઆɾඪ ؆୯ʹϑΣΠΫχϡʔεఆʹ৮ΕΒΕΔ ڥ͕ඞཁͰ͋ΔͷͰͳ͍͔ɻ
੍࡞ ✌੍࡞
IUUQTUXJUUFSDPNUSFOE@PXM
σβΠϯͷϙΠϯτ ը૾ʹΛॻ͔ͳ͍ ҙࣄ߲ΛಡΜͰΒ͏ͨΊ ʹɺ͋͑ͯςΩετʹ͔݁͠ ՌΛॻ͖·ͤΜͰͨ͠ πΠʔτΛݟΔͱ͖ʹ ਓؒը૾͔ΒݟΔ 👎
੍࡞ 👩🎓ੳख๏
Ϟσϧºػցֶश Ҿ༻ɿIUUQTXXXSJLFMBCKQTUVEZ ਖ਼͍͠ใ '",&
IUUQTXXXBBBJPSHPDTJOEFYQIQ"""*"""*QBQFSWJFX'JMF அख๏ ɾΞΧϯτ͕ొ͞Ε͔ͯΒͲΕ͙Β͍͔ ɾݩπΠʔτ͔πΠʔτ͞Ε͔ͯΒͲΕ͙Β͍ܦա͍ͯ͠Δ͔ ɾೝূࡁΈΞΧϯτ͔Ͳ͏͔ ɾϓϩϑΟʔϧը૾Λࢦఆ͍ͯ͠Δ͔Ͳ͏͔ ɾ͍͍Ͷ͍ͯ͠Δ
ɾϑΥϩϫʔͷ ɾϑΥϩʔ͍ͯ͠Δ ɾ໊લͷจࣈ ɾࣗݾհͷจࣈ πΠʔτจষʹҰͱΒΘΕͣੳΛߦ͏ 📌 ࢀߟจ
੍࡞ ✍Ϟσϧͷਫ਼ɾಁ໌ੑ
ֶशϞσϧਫ਼͑ Ϟσϧͷ༧ଌ ࣮ ࡍ
ಁ໌ੑϨϙʔτΛ࡞ެ։ tensorboardXΛ༻͍ͨϞσϧͷՄࢹԽ
ੳ 🧠 πΠʔτऩू IUUQTUXJUUFSDPNUSFOE@PXM ϑΣΠΫੳ 🤖 ใࠂ
IUUQTUSFOEPXMTVHPLVOBSJUBJEFW ϙʔλϧαΠτ աڈͷੳ͕ݟΕΔ ੳϦΫΤετ͕ૹΕΔ
"NB[PO$MPVE8BUDI&WFOU πΠʔτ "NB[PO4 "NB[PO-BNCEB "NB[PO%ZOBNP%# ఆظ࣮ߦ ੳ ཧը໘ ϩά ը૾Ξοϓϩʔυ
ٕज़ߏ (PPHMF"QQ4DSJQU ϑϩϯτΤϯυ
੍࡞ 😁5SFOE0XMͷظ
ϑΣΠΫ χ ϡ ʔε ૣ ͘ ͕ Δ
ιʔείʔυϞσϧɺݟͷެ։ ڵຯΛ࣋ͬͯͩ͘͞ΔํΛ૿͠ ք۾શମͷٕज़ྗΛ্͛Δ ࠓޙͷల ҙݟΛΒ͏ɾΊΔ ݚڀػؔاۀͱͷڞಉݚڀ ༷ʑੈʹ֦ࢄ
ެ։ɾڞ༗͢Δ ਫ਼ΛߴΊΔ ࣗવݴޠॲཧͱϞσϧͷྑΛ ༻͍ͨϞσϧͷ։ൃ ΞϓϩʔνΛ૿͢ 5XJUUFS8FCҎ֎ͰͷϦϦʔε ༷ʑͳΞόλʔઃఆͰݕূΛߦ͏
None
༧උࢿྉ
ϑϩϯτΤϯυ Ϧʔμʔ ػցֶशɾαʔόʔαΠυ Α͏͔Μ ͿΜͿΜ ͍ʔͪΌΜ 5ZMFS Ωϯδϣ ͱΈͬ͘͢ αʔϕΠɾσβΠϯ
αʔϕΠɾϢʔβΠϯλϏϡʔ αʔόʔαΠυ
4DSBQCPY νʔϜ࿈ܞ /PUJPO 4MBDL
ͳͥɺʮϑΝΫτνΣοΫʯͱ͍͏ ΞϓϩʔνΛऔ͍ͬͯΔͷ͔ʁ
ใੜଶܥਐԽ͠ଓ͚͍ͯΔ Ҿ༻ɿIUUQTXXXEFMPJUUFDPNKQKBQBHFTTUSBUFHZBSUJDMFTDCTJOGPSNBUJPOFQJEFNJDIUNM ʢੈلͰສഒʹ૿େͨ͠ใୡྗʙใͷٸͳછʮΠϯϑΥσϛοΫʯͱʣ
None
օ͞ΜͷྗΛି͍ͯͩ͘͠͞ 5XJ UU FS ͷதͷਓͱ͓Γ߹͍ͷํ ޙ΄Ͳ͓ܨ͍͖͍͗ͨͩͨͰ͢ʂ
ࢲୡ͕͍ͬͯΔ͜ͱԿ͕ຊൃɺੈքॳͷϓϩμΫτͳͷ͔ ैདྷڝ߹ͱԿ͕ҧ͏ͷ طଘͷΈͱͷҧ͍ /-1 ࣗવݴޠॲཧ ͰͰ͖Δ͜ͱ طଘͷࣙॻɾίʔύεΛͱʹ จ຺Λղੳ͢Δ ٯʹݴ͑ ৽͍͠ࣄɾະͷʹ
ରॲͰ͖ͳ͍͔͠Εͳ͍
ࢲୡ͕͍ͬͯΔ͜ͱԿ͕ຊൃɺੈքॳͷϓϩμΫτͳͷ͔ ैདྷڝ߹ͱԿ͕ҧ͏ͷ طଘͷΈͱͷҧ͍ Ϟσϧղܾ͠·͢ɻ ະͷࣄ࣮ʹରͯ͠ ٬؍తͳੳ͕Մೳ
ࢲୡ͕͍ͬͯΔ͜ͱԿ͕ຊൃɺੈքॳͷϓϩμΫτͳͷ͔ ैདྷڝ߹ͱԿ͕ҧ͏ͷ ػցֶशϞσϧͷΈ 3// -45. ࠶ؼܕχϡʔϥϧ ωοτϫʔΫ ϩʔΧϧͳੳ $//
ΈࠐΈχϡʔϥϧ ωοτϫʔΫ άϩʔόϧͳੳ /-1Ͱকདྷతʹਫ਼্Λʂ [ࢀߟจݙ] Yang Liu, Yi-Fang Brook Wu (2018) Early Detection of Fake News on Social Media Through Propagation Path Classi fi cation with Recurrent and Convolutional Networks