Upgrade to Pro — share decks privately, control downloads, hide ads and more …

さあ、異常を検知しよう！！

Sponsored · SiteGround - Reliable hosting with speed, security, and support you can count on. →
Avatar for tsurubee tsurubee
March 23, 2018
Technology
470
0

 さあ、異常を検知しよう！！

異常検知の話

Avatar for tsurubee

tsurubee

March 23, 2018

More Decks by tsurubee

See All by tsurubee
大規模言語モデルにおけるData-Centric AIと合成データの活用 / Data-Centric AI and Synthetic Data in Large Language Models
Avatar for tsurubee tsurubee
1
570
言語モデルによるAI創薬の進展 / Advancements in AI-Driven Drug Discovery Using Language Models
Avatar for tsurubee tsurubee
2
610
AIトップカンファレンスからみるData-Centric AIの研究動向 / Research Trends in Data-Centric AI: Insights from Top AI Conferences
Avatar for tsurubee tsurubee
3
3.3k
DeepCrysTet: A Deep Learning Approach Using Tetrahedral Mesh for Predicting Properties of Crystalline Materials
Avatar for tsurubee tsurubee
0
1.3k
3次元メッシュで表現した結晶構造を用いた材料物性の予測に向けた深層学習モデルの設計 / Design of Deep Learning Model for Predicting Material Properties Using Crystal Structure Represented by Three-Dimensional Mesh
Avatar for tsurubee tsurubee
1
2.5k
分散システムの性能異常に対する機械学習の解釈性に基づく原因診断手法 / A Method for Diagnosing the Causes of Performance Issues in Distributed Systems Based on the Interpretability of Machine Learning
Avatar for tsurubee tsurubee
0
1.7k
機械学習の解釈性に関する研究動向とシステム運用への応用 / A Survey on Interpretable Machine Learning and Its Application for System Operation
Avatar for tsurubee tsurubee
0
390
機械学習モデルの局所的な解釈に着目したシステムにおける異常の原因診断手法の構想
Avatar for tsurubee tsurubee
0
8.1k
アニーリングマシンを活用したエッジAIにおける
生成モデルの学習効率化のためのアーキテクチャ
Avatar for tsurubee tsurubee
0
1.7k

Other Decks in Technology

See All in Technology
現地で盛り上がった WWDC26 Keynote
Avatar for ZOZO Developers zozotech
PRO
1
210
小さく始める AI 活用推進 ― 日経電子版 Web チームの事例/nikkei-tech-talk47
Avatar for 日本経済新聞社 エンジニア採用事務局 nikkei_engineer_recruiting
0
230
EventBridge Connection
Avatar for kensh _kensh
5
700
AIはどのように 組織のアジリティを変えるのか？
Avatar for J.K junki
1
470
タクシーアプリ『GO』の実践的データ活用
Avatar for GO Inc. dev mot_techtalk
3
190
Kubernetesにおける学習基盤とLLMOpsの概要
Avatar for ry ry
1
250
2026TECHFRESH畢業分享會 - Lightning Talk - E起 See See : 電商推薦讀心術? 數據說了算
Avatar for LINE Developers Taiwan line_developers_tw
PRO
0
850
中期計画、2回作ってみた　～業務委託と正社員、両方の視点から～
Avatar for 株式会社出前館 demaecan
1
690
2026TECHFRESH畢業分享會 - Lightning Talk - 資料也要 CI/CD? 用 Airbyte 自動化資料同步
Avatar for LINE Developers Taiwan line_developers_tw
PRO
0
850
AWSシリコン最前線 〜AI時代のチップ選択を読み解く〜
Avatar for Hiroshi Tokoyo htokoyo
2
500
白金鉱業Meetup_Vol.24_「AIエージェントは分けるほど良い」は本当か？ / Is it true that “the more you divide AI agents, the better”?
Avatar for BrainPad brainpadpr
1
310
Dario Amodi『Policy on the AI Exponential』を理解する
Avatar for 長津孝輔 nagatsu
0
230

Featured

See All Featured
JavaScript: Past, Present, and Future - NDC Porto 2020
Avatar for David Neal reverentgeek
52
6k
世界の人気アプリ100個を分析して見えたペイウォール設計の心得
Avatar for Akihiro Kokubo akihiro_kokubo
PRO
71
40k
Groundhog Day: Seeking Process in Gaming for Health
Avatar for Sebastian Deterding codingconduct
0
200
The Invisible Side of Design
Avatar for Vitaly Friedman smashingmag
302
52k
Digital Projects Gone Horribly Wrong (And the UX Pros Who Still Save the Day) - Dean Schuster
Avatar for UX Y'all uxyall
1
1.7k
YesSQL, Process and Tooling at Scale
Avatar for Rocio Delgado rocio
174
15k
Data-driven link building: lessons from a $708K investment (BrightonSEO talk)
Avatar for Szymon Słowik szymonslowik
1
1.1k
Heart Work Chapter 1 - Part 1
Avatar for Lake Fama lfama
PRO
7
36k
Hiding What from Whom? A Critical Review of the History of Programming languages for Music
Avatar for Tomoya Matsuura tomoyanonymous
2
850
Leadership Guide Workshop - DevTernity 2021
Avatar for David Neal reverentgeek
1
300
StorybookのUI Testing Handbookを読んだ
Avatar for Shingo Yamazaki zakiyama
31
6.8k
Sharpening the Axe: The Primacy of Toolmaking
Avatar for Bryan Cantrill bcantrill
46
2.9k

Transcript

  1. !UTVSVCFF(.01FQBCP *OD 5FDI.5( ͋͞ɺҟৗΛݕ஌͠Α͏ʂʂ

  2. ΤϯδχΞ !UTVSVCFF ϗεςΟϯάࣄۀ෦ΠϯϑϥνʔϜ ݄೔ʹೖࣾ͠·ͨ͠ʂ

  3. ࠓճ͸ɾɾ

  4. Πϯϑϥ͡Όͳͯ͘ ҟৗݕ஌ͷ࿩

  5. ҟৗݕ஌ͱ͸ େଟ਺ͷσʔλͱ͸ৼΔ෣͍͕ ҟͳΔσʔλΛݕग़͢Δٕज़ σʔλϚΠχϯά نଇੑɹҟৗ σʔλͷࢁ

  6. օ͞ΜɺҟৗΛݕ஌͍ͨ͠ͱ͖ ͋Γ·͢ΑͶʁʁ

  7. ҟৗݕ஌ͷԠ༻ઌ ίϯϐϡʔλ΢Πϧε΍%PT߈ܸͷૣظൃݟ ηΩϡϦςΟ෼໺ ނো༧ஹݕ஌ ػց෼໺ ྲྀߦͷݕ஌ɾ৽τϐοΫͷൃݟɾ ϢʔβߦಈͷมԽݕ஌ ϚʔέςΟϯά෼໺

  8. Πϯϑϥͱҟৗݕ஌ IUUQTXXXDPNQVUFSPSHDTEMUSBOTULLIUNM .4#MBTUͷݕ஌

  9. IUUQTTQFBLFSEFDLDPNTVHJZBNBNBDLFSFMNFFUVQOVNCFS Πϯϑϥͱҟৗݕ஌ .BDLFSFMͷҟৗݕ஌

  10. ϖύϘͰ΋ɾɾ IUUQTTQFBLFSEFDLDPNNBUTVNPUP@SJPU IUUQTTQFBLFSEFDLDPNNPOPDISPNFHBOFTDIFEVMFEBVUPTDBMJOHPGWJSUVBMTFSWFSTCZBDDFTTGSFRVFODZQSFEJDUJPO

  11. ࠓ೔͸؆୯ͳྫΛ΋ͱʹ ҟৗݕ஌ͷงғؾΛ௫Ή ʢ਺ࣜ͸ग़͠·ͤΜʣ

  12. ҟৗσʔλͷྫ %BWJTσʔλɿਓͷੑผɺଌఆͨ͠਎௕ɾମॏͷσʔλ IUUQTWJODFOUBSFMCVOEPDLHJUIVCJP3EBUBTFUTEBUBTFUTIUNM

  13. ҟৗσʔλͷྫ %BWJTσʔλɿਓͷੑผɺଌఆͨ͠਎௕ɾମॏͷσʔλ IUUQTWJODFOUBSFMCVOEPDLHJUIVCJP3EBUBTFUTEBUBTFUTIUNM ҟৗʂ

  14. ҟৗσʔλͷྫ %BWJTσʔλɿਓͷੑผɺଌఆͨ͠਎௕ɾମॏͷσʔλ IUUQTWJODFOUBSFMCVOEPDLHJUIVCJP3EBUBTFUTEBUBTFUTIUNM ҟৗʂ if value > 120: print('ERROR!') JGจͰݕ஌Ͱ͖ͦ͏ʂ

  15. Ұ൪γϯϓϧ͔ͭϙϐϡϥʔͳ ҟৗݕ஌ख๏ɿᮢ஋ઃఆ

  16. ᮢ஋ઃఆͷ໰୊఺ᶃ ͷࠜڌ͸Կʁ໰୊ ҟৗʂ if value > 120: print('ERROR!') JGจͰݕ஌Ͱ͖ͦ͏ʂ

  17. ϗςϦϯάཧ࿦ ͪΌΜͱ΍Δʹ͸౷ܭֶ ֎Ε஋ ਖ਼ن෼෍ΛԾఆ͠ɺσʔλੜ੒ͷ֬཰ϞσϧΛֶश͢Δ ౷ܭֶΛ࢖ͬͯ٬؍తͳਫ४Ͱݕ஌͢Δ ֎Ε஋

  18. ᮢ஋ઃఆͷ໰୊఺ᶄ ࣌ܥྻมԽʹରԠͰ͖ͳ͍໰୊ ࣌ؒ ͬͪ͜͸0, ᮢ஋ม͑ͳ͍ͱʂ อकͭΒ͍ɾɾ ࣌ؒͷܦաͱͱ΋ʹ্ঢ܏޲ʹ͋Δ ࣌ؒͷܦաʹແؔ܎ ԣ࣠͸࣌ؒͰ͢ʂ

  19. ࣌ܥྻมԽʹͲ͏΍ͬͯ ରԠ͠Α͏

  20. Ұ൪γϯϓϧͳΞϓϩʔν

  21. ೖྗσʔλΛͣΒͯ͠΍ΔͱউखʹରԠͯ͘͠ΕΔ εϥΠυ૭ ͜ͷதͰฏۉΛͱΔ ฏۉ஋ ฏۉ͔Βͷ ڑ཭͕ҟৗ౓

  22. ͜ΕΛൃల͍ͤͯ͘͞ͱɾɾ

  23. 1ZUIPO੡ҟৗݕ஌ϥΠϒϥϦ#BOQFJ

  24. ಛҟεϖΫτϧม׵ ࣌ࠁ શମΠϝʔδ ʮೖ໳ػցֶशʹΑΔҟৗݕ஌ ʕ3ʹΑΔ࣮ફΨΠυʯͷਤ Λࢀߟʹ࡞੒ աڈଆͱݱࡏଆͷߦྻಉ࢜ͷ৯͍ҧ͍ͷେ͖͞ΛఆྔԽ ཤྺߦྻ9 ςετߦྻ9 ओ෦෼ۭؒ\V

    VN ^ \R RN ^ มԽ౓

  25. σϞ

  26. #BOQFJʹΑΔҟৗݕ஌ ಛҟεϖΫτϧม׵ʹΑΔ܏͖ͷมԽ఺ݕ஌ IUUQTXXXZPVUVCFDPNXBUDI W@XPVC-"I9L ಛҟεϖΫτϧม׵ʹΑΔप೾਺ҟৗݕ஌ IUUQTXXXZPVUVCFDPNXBUDI WF'7/,"7U/1

  27. ·ͱΊ • ҟৗݕ஌͸Πϯϑϥʹ΋Ԡ༻Ͱ͖Δ • Ұ൪γϯϓϧͳҟৗݕ஌͸ᮢ஋ઃఆɺͰ΋໰୊΋͋Δ • ٬؍తͳਫ४Ͱҟৗݕ஌͢Δʹ͸౷ܭֶ • ࣌ܥྻͳมԽʹ͸εϥΠυ૭ͳͲͷςΫχοΫ͕࢖͑Δ

  28. ͝ਗ਼ௌ͋Γ͕ͱ͏͍͟͝·ͨ͠ʂ