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
スマートフォン向けインターネット広告配信システムの配信最適化
Search
Sponsored
·
Ship Features Fearlessly
Turn features on and off without deploys. Used by thousands of Ruby developers.
→
Takashi Nishibayashi
July 11, 2017
Business
2k
3
Share
スマートフォン向けインターネット広告配信システムの配信最適化
DATUM STUDIO Conference 2017夏での講演資料です
非エンジニア向けの内容です
Takashi Nishibayashi
July 11, 2017
More Decks by Takashi Nishibayashi
See All by Takashi Nishibayashi
病院向け生成AIプロダクト開発の実践と課題
hagino3000
0
620
入院医療費算定業務をAIで支援する:包括医療費支払い制度とDPCコーディング (公開版)
hagino3000
0
210
診断前の病歴テキストを対象としたLLMによるエンティティリンキング精度検証
hagino3000
1
190
論文紹介 Improving Medical Reasoning through Retrieval and Self-Reflection with Retrieval-Augmented Large Language Models
hagino3000
0
960
論文紹介 Audience Size Forecasting Fast and Smart Budget Planning for Media Buyers
hagino3000
0
260
論文紹介 Towards a Fair Marketplace: Counterfactual Evaluation of the trade-off between Relevance, Fairness & Satisfaction in Recommendation Systems
hagino3000
1
670
論文紹介 Budget Management Strategies in Repeated Auctions (公開版)
hagino3000
2
330
論文紹介 A Request-level Guaranteed Delivery Advertising Planning: Forecasting and Allocation
hagino3000
1
160
論文紹介 Online Experimentation with Surrogate Metrics Guidelines and a Case Study
hagino3000
1
420
Other Decks in Business
See All in Business
Remote First: Building Distributed Teams that Win
jnunemaker
PRO
1
130
株式会社ELYZA(イライザ) 採用情報資料 / RECRUIT PITCH
elyza
2
750k
1+1=3になる世界
matsuzaka01
0
160
FABRIC TOKYO会社紹介資料 / We are hiring(2026年04月02日更新)
yuichirom
36
390k
Canary Inc. Company Deck
canaryinc
PRO
2
16k
【ラクス】新卒採用
rakus_career
0
75k
20260422_PdM社内定例資料_スタンスをとることがアウトカムに直結する
numashi
4
2.4k
AnyMind Group Credential Deck(EN)
anymind
3
81k
ele&company_companydeck
eleand
0
4.6k
株式会社SAFELY 会社紹介 / Company
safely_pr
1
6.6k
CX Lens 購入後体験(ポストパーチェス)分析レポート
contentmetrics
0
140
Tableau巣鴨会_20260328_LT
hayashi_ds017
0
200
Featured
See All Featured
Leading Effective Engineering Teams in the AI Era
addyosmani
9
1.9k
brightonSEO & MeasureFest 2025 - Christian Goodrich - Winning strategies for Black Friday CRO & PPC
cargoodrich
3
680
What the history of the web can teach us about the future of AI
inesmontani
PRO
1
530
Bash Introduction
62gerente
615
210k
JavaScript: Past, Present, and Future - NDC Porto 2020
reverentgeek
52
5.9k
We Have a Design System, Now What?
morganepeng
55
8.1k
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
akihiro_kokubo
PRO
70
39k
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
21
1.5k
Information Architects: The Missing Link in Design Systems
soysaucechin
0
900
VelocityConf: Rendering Performance Case Studies
addyosmani
333
25k
How GitHub (no longer) Works
holman
316
150k
RailsConf 2023
tenderlove
30
1.4k
Transcript
εϚʔτϑΥϯ͚ Πϯλʔωοτࠂ৴γεςϜ ʹ͓͚Δ৴࠷దԽ 5BLBTIJ/JTIJCBZBTIJ %"56.456%*0$POGFSFODFՆ
Agenda 1.ࣗݾհ 2.ΞυωοτϫʔΫͱωοτࠂۀքʹ͍ͭͯ 1.ωοτࠂϏδωεͷ֓ཁ 2.ࠂ৴ͰΘΕΔਓೳཁૉٕज़ 3.ฐࣾࣄྫͷհ 1.ΫϦοΫ୯Ձͷௐઅ 2.৴͢Δࠂͷબ
ࣗݾհ ID: hagino3000 Name: ྛ (Takashi Nishibayashi) Job: Software
Engineer ݱࡏZucks AdNetworkʹͯ৴ϩδοΫ ͷ։ൃΛ୲ɻσʔλੳج൫ͷߏங͔Β ػցֶशΛͬͨ༧ଌɺ࠷దԽॲཧ·Ͱɻ
࠷ۙͷ׆ಈ ਓೳֶձࢽ Vol. 32 No. 4 (2017/07) ͷʮࠂͱAI ಛूʯʹຊͷ༰ʹؔ࿈ ͨ͠هࣄΛدߘ͍ͯ͠·͢ɻ
ৄࡉʹڵຯ͕͋Δํ͝Ұಡ͍ͩ͘͞ɻ
Ad Networkͱ ✴ ΠϯλʔωοτͷσΟεϓϨΠࠂྖҬʹ͓͍ͯɺ ෳͷࠂओͱෳͷഔମࣾΛଋͶͯࠂΛ৴͢ ΔΈ ✴ ഔମࣾʹऩӹΛɺࠂओʹίϯόʔδϣϯΛ ͨΒ͢ͷ͕ࣄ ✴
ࠂϦΫΤετຖʹͲͷࠂΛ৴͢Δ͔ϩδοΫ Ͱܾఆ͍ͯ͠Δ
Ad Network ࠂೖߘ ࠂओ ഔମࣾ ࠂഔମ (ϝσΟΞ) ࠂ৴ ࠂඅ
ࠂऩӹ ΦʔσΟΤϯε Click
ωοτࠂ৴ͱ ਓೳཁૉٕज़ ✴ ৴͢Δࠂͷબ ✴ CTRɾCVR༧ଌ ✴ ϢʔβʔτϥοΩϯά ✴ ྫ:
ෳσόΠεΛލ͍ͩߪങߦಈͷ ✴ ࠂޮՌͷਪఆ ✴ ྫ: TV CMͷޮՌਪఆ ✴ ࠂΦʔΫγϣϯʹ͓͚ΔϦΞϧλΠϜೖࡳ
Zucks AdNetworkʹ͓͚Δ ࣄྫհ ✴ લఏ ✴ ৫ͷσʔλ׆༻εςʔδ ✴ Ad NetworkʹٻΊΒΕΔ৴
✴ ࣄྫ1. ΫϦοΫ୯Ձͷ࠷దԽ ✴ ࣄྫ2. ୳ࡧ৴ͷޮԽ
৫ͷσʔλ׆༻εςʔδ σʔλΛཷΊΒΕΔ σʔλ͕ར༻Ͱ͖ͳ͍ॴʹ͍͖ͳΓػցֶशΛͬͨ γεςϜΛσϓϩΠͰ͖ͳ͍ σʔλ͕Ҿ͖ग़ͤΔ ੳ͕Ͱ͖Δ ༧ଌॲཧͷγεςϜԽ ༧ଌ݁ՌΛͬͨऩӹͷ࠷େԽ ݕূͷ
Έ #*πʔϧͷಋೖ "#ςετ ҼՌޮՌਪ ཧ࠷దԽ ػցֶश ੳج൫ͷߏங
ཁһ֬อ ✴ ࠷ॳ͔ΒશͯͷϨΠϠʔͰඞཁͳεΩϧΛ࣋ͭਓࡐ Λἧ͑Δͷ͍͠ ✴ Γͳ͍ॴ͍͍ײ͡ʹิ͍ͬͯ͘ඞཁ͕͋Δ ✴ ֎෦……??
ղ͖͍ͨ ✴ ͍ͭ ✴ ୭ʹ or Ͳͷࠂʹ ✴ ͲͷࠂΛ ✴
(ΫϦοΫ୯Ձ) ͍͘ΒͰ ✴ දࣔ͢Δͷ͔
ఆࣜԽ ✴ ඪ ✴ ഔମࣾऩӹͷ࠷େԽ ✴ ੍݅ ✴ ࠂओͷඪCPA (ίϯόʔδϣϯ֫ಘ͋ͨΓͷίετ)
✴ ࠂओͷ༧ࢉ ✴ ࠂදࣔճ ͨͩ͠ ΫϦοΫɾίϯόʔδϣϯ ৴͠ͳ͍ͱΘ͔Βͳ͍
Ұͭͷ࠷దԽͱͯ͠ղ͚Εྑ͍ ͷͰ͕͢ɺ͍͠ͷͰෳͷʹ ͚ͯ։ൃͯ͠·͢
ΫϦοΫ୯Ձͷௐ ✴ CPA (ίϯόʔδϣϯ͋ͨΓͷ֫ಘίετ) Λࠂओ ͷཁʹ߹ΘͤΔͷ͕త ✴ ͋ΔࠂΩϟϯϖʔϯΛ৴͢Δͱͯ͠ ✴ ίϯόʔδϣϯ͕औΕΔͷ୯Ձ্͍͛ͨ
՝ ✴ ྫ ✴ ίϯόʔδϣϯ100% ✴ ΫϦοΫ୯Ձ100ԁͳΒCPA100ԁͱͳΔ ✴ ͳΔ͘৴ͷॳظஈ֊ʹ͓͍ͯίϯόʔδϣϯ Λਪఆ͍ͨ͠
✴ ͔͠͠ɺ৴ॳظΫϦοΫͷαϯϓϧαΠζ͕খ ͘͞౷ܭతʹྑ͍ͱѱ͍ͱݴ͑ͳ͍
CVRਪఆ ✴ ίϯόʔδϣϯͷࣅͨಉ࢜Ͱ͋ΕɺCVRۙ͘ ͳΔͣɻ͜ΕΛࣄલͱͯ͑͠ͳ͍͔ ✴ ࣅͨಉ࢜ͷू߹ΫϥελϦϯάͰٻΊΔ ✴ ࣄલ֬Λಋೖ͠ɺϕΠζͷఆཧʹΑΓΫϦοΫ n ͷ
͏ͪ k ݸͷίϯόʔδϣϯΛ؍ଌͨ͠ޙͷ CVR ͷࣄޙ ֬Λߟ͑ΔɻCVRͷࣄલΛϕʔλBeta(α, β) ͱ͢ΔͱɺCVRͷࣄޙϕʔλʹͳΔɻ
݁Ռݕূ1 ✴ ༧ଌਫ਼ΦϑϥΠϯ࣮ݧͰݕূͰ͖Δ ✴ RMSE, Accuracy, Precision, F-value …. ✴
ϏδωεαΠυ͕Γ͍ͨͷɺ༧ଌ͕ͨΔࣄʹΑ ΔܦӦࢦඪͷӨڹ (ྫ: ച্) ༧ଌਫ਼͕YY্͕Γ·ͨ͠ ച্Ͳ͏ͳΔͷʜʜ
݁Ռݕূ2 ✴ ࣮ࡍʹCPA͕ඪCPAʹۙ͘ͳΔͷ͔ɺຊ൪ʹϦϦʔ εͯ͠ݕূ ✴ log(࣮CPA/ඪCPA) ΛطଘϩδοΫద༻Ωϟϯ ϖʔϯͱൺֱɻରͰݟΔͷɺ2ഒʹͳΔͷͱ ʹͳΔͷΛಉ͡ΠϯύΫτͱͯ͠ଊ͑ΔͨΊɻ ✴
log(࣮CPA/ඪCPA) ͷʹ͍ͭͯϊϯύϥϝτ ϦοΫݕఆͰ͕ࠩ͋Δࣄͷ֬ೝͱ4ҐͷࠩΛΈΔ
ެ։൛ʹ͖ͭআ ݁Ռ
৴͢Δࠂͷબ ✴ ഔମऀऩӹͷߴ͍ࠂΛଟ͘৴͍ͨ͠ ✴ ΫϦοΫ͕ଟ͘ίϯόʔδϣϯऔΕΔ ✴ ݁ՌతʹΫϦοΫ୯Ձ্͛ΒΕΔ ✴ ޮՌ͕ྑ͍͔ѱ͍͔৴͠ͳ͍ͱΘ͔Βͳ͍ ✴
ࠂͱࠂͷΈ߹Θͤແʹ͋ΔͷͰૣ͘ྑ ͍Έ߹ΘͤΛҾ͖͍ͯͨ ✴ ࣝͷ׆༻ͱ୳ࡧͷδϨϯϚ
୳ࡧͱ׆༻ ✴ ׆༻ ✴ ʹͱͬͯऩӹ͕ߴ͍ͱΘ͔͍ͬͯΔࠂΛ৴ ✴ ࠷ߴ͍ͷΈΛ৴ͨ͠Βྑ͍༁Ͱͳ͍ ✴ طଘͷόϯσΟοτΞϧΰϦζϜΛͦͷ··͍ ʹ͍͘
✴ ୳ࡧ ✴ ʹͱͬͯऩӹ͕ະͷࠂΛ৴͢Δ
ଟόϯσΟοτʹΑΔ Ξϓϩʔν ✴ εϩοτϚγϯͷϝλϑΝʔ ✴ εϩοτϚγϯ͕ෳ͋ͬͨ࣌ʹͲΕΛԿճҾ͘ ͖͔ ✴ ϝϦοτ ✴
ڭࢣσʔλ͕ແ͍ॴ͔ΒελʔτͰ͖Δ ✴ ৽͍͠ࠂΩϟϯϖʔϯ͕࣍ʑͱೖߘ͞ΕΔઃఆ
୳ࡧͷޮԽ ✴ ͋Δʹ͓͚Δɺଞͷࠂͱͷൺֱ ✴ ଞͷࠂͱൺֱͯ͠ऩӹ͕ѱ͍ͱΘ͔ͬͨ࣌Ͱ୳ ࡧΛΊΕྑ͍ ✴ ऩӹੑ(eCPM)ͷ্քΛ͏ ✴ ֬తʹߴͯ͘͜Ε͙Β͍ͩΖ͏ɺͱ͍͏
✴ ৴Λଓ͚ΔࣄͰԼ͕͍ͬͯ͘
ݕূ ✴ ͷऩӹ͕Ͳ͏มԽ͔ͨ͠Λݟ͍ͨ ✴ ͔͠͠ɺऩӹ࣌ؒมԽͷӨڹΛڧ͘ड͚Δ ✴ ظʹ༧ࢉফԽ͕͋ΔͨΊɺඞͣ৳ͼΔ ✴ ୯७ʹϩδοΫมߋલޙͰൺֱͰ͖ͳ͍
ϥϯμϜԽൺֱࢼݧʹΑΔݕূ ϩδοΫͷมߋʹΑΔհೖ σʔλαϯϓϧɺԣ࣠࣌ؒ ࠂΛ܈ʹ͚ͯσʔλΛऔΔ
݁Ռ ެ։൛ʹ͖ͭআ
͋Γ͕ͱ͏͍͟͝·ͨ͠