Giter Club home page Giter Club logo

jwj51720 / kdis-datascience-contest Goto Github PK

View Code? Open in Web Editor NEW
0.0 1.0 0.0 18.54 MB

๐Ÿฅ‡ ๋Œ€์ƒ(1์œ„) 2023 KDI๊ตญ์ œ์ •์ฑ…๋Œ€ํ•™์› ๋ฐ์ดํ„ฐ์‚ฌ์ด์–ธ์Šค ํ•™๋ถ€์ƒ ๊ณต๋ชจ์ „

Home Page: https://enchanted-gum-e41.notion.site/KUBIG-2023-2-Conference-Project-ad83c4aa06354c59aa12761c020ddeec?pvs=4

Jupyter Notebook 99.90% R 0.10%
data-analysis political-data-analysis

kdis-datascience-contest's Introduction

๐ŸŒ Proposals for the Future Direction of South Koreaโ€™s Official Development Assistance

2023.09.30 - 2023.12.01
๐Ÿฅ‡ ๋Œ€์ƒ(1์œ„) ์ˆ˜์ƒ

๐Ÿ“’Team Notion

์ฒจ๋ถ€ 1  KDIS ๊ณต๋ชจ์ „_ํ•œ๊ตญ์˜ ๋ฏธ๋ž˜ ODA ์‹œํ–‰๋ฐฉํ–ฅ์— ๋Œ€ํ•œ ์ œ์–ธ_1

Abstract

  • ํ•œ๊ตญ์€ 1995๋…„ ์„ธ๊ณ„์€ํ–‰์˜ ์›์กฐ๋Œ€์ƒ๊ตญ์„ ๋ฒ—์–ด๋‚˜ 2009๋…„ DAC ํšŒ์›๊ตญ์ด ๋˜์–ด ODA(Official development assistance, ๊ณต์ ๊ฐœ๋ฐœ์›์กฐ)๋ฅผ ์ง„ํ–‰
  • ํ•œ๊ตญ์˜ ODA ๊ทœ๋ชจ๋Š” 2010๋…„ 10์–ต ๋‹ฌ๋Ÿฌ ๊ทœ๋ชจ๋ฅผ ๋„˜์–ด์„œ 2016๋…„ 20์–ต ๋‹ฌ๋Ÿฌ ๊ทœ๋ชจ๋ฅผ ์ดˆ์›”
  • ์–ธ๋ก  ๋ณด๋„์— ๋”ฐ๋ฅด๋ฉด ํ•œ๊ตญ์˜ ODA ๊ทœ๋ชจ๋Š” ๋งค๋…„ ์ฆ๊ฐ€ํ•˜๋Š” ์ถ”์„ธ์ด๋‚˜ ๊ทœ๋ชจ๋งŒ์œผ๋กœ ODA์˜ ์‹ค์ ์„ ํ‰๊ฐ€ํ•  ์ˆ˜๋Š” ์—†์Œ
    • ๊ตญ๋ณ„ ํ”„๋กœ๊ทธ๋žจ ์›์กฐ๋Š” 2010๋…„ 54.3%์—์„œ 2021๋…„ 47.5%๋กœ ๊ฐ์†Œ
    • CPA ๋น„์ค‘์ด ๋‹ค๋ฅธ ํšŒ์›๊ตญ์— ๋น„ํ•ด ํฌ๊ฒŒ ๋‚ฎ์•„ ODA์˜ ํšจ์œจ์„ฑ ์ธก๋ฉด์—์„œ ์ตœํ•˜์œ„ ํ‰๊ฐ€
    • ํ•œ๊ตญ์˜ ๊ฒฝ์ œ ์ง€ํ‘œ(GNI, ๋ฌผ๊ฐ€์ƒ์Šน๋ฅ , ๊ฒฝ์ œ์„ฑ์žฅ๋ฅ )์„ ๊ณ ๋ คํ–ˆ์„ ๋•Œ ์‹ค์งˆ์  ์ธก๋ฉด์—์„œ์˜ ์ฆ์•ก ์—ฌ๋ถ€๋Š” ๋ถˆํˆฌ๋ช…
    • ๋น„๊ต์  ๋‚ฎ์€ ๋ฌด์ƒ์›์กฐ ๋น„์œจ, ์ •์ฑ…/์ง‘ํ–‰ ๋‹ด๋‹น ๊ธฐ๊ด€ ๋ถ„๋ฆฌ๋กœ ์ธํ•œ ์›์กฐ ๋ถ„์ ˆํ™”, ์˜ˆ๋น„๊ฒ€ํ† ์ œ์— ๊ธฐ์ธํ•œ ๋Š๋ฆฐ ์ •์ฑ… ์ถ”์ง„ ๋“ฑ ์—ฌ๋Ÿฌ ์ •์ฑ…์  ๋ฌธ์ œ
    • ์‹ค์ œ๋กœ UN์˜ ์„ ์ง„๊ตญ ODA/GNI(๊ตญ๋ฏผ์ด์†Œ๋“ ๋Œ€๋น„ ์›์กฐ์•ก) ๋ชฉํ‘œ์น˜์ธ 0.7%์— ๋น„ํ•ด ํ•œ๊ตญ์€ 2022๋…„ 0.17%๋กœ 30๊ฐœ ํšŒ์›๊ตญ ์ค‘ 16์œ„
  • ์ด์— ODA ์ƒํ™ฉ ๊ฐœ์„ , ์˜ˆ์‚ฐ์˜ ํšจ์œจ์ ์ธ ์ง‘ํ–‰ ๋ฐ ODA์˜ ํšจ์šฉ ๊ทน๋Œ€ํ™”๋ฅผ ์œ„ํ•œ ๋ฐ์ดํ„ฐ ๊ธฐ๋ฐ˜์˜ ์‹ฌ์ธต์  ๋ถ„์„ ์ˆ˜ํ–‰
    • ๊ตญ๋‚ด์™ธ ODA์™€ ๊ด€๋ จํ•œ ์—ฌ๋Ÿฌ ๋ถ„์•ผ์˜ ์ •๋Ÿ‰์  ์ง€ํ‘œ ์ˆ˜์ง‘, ์ „์ฒ˜๋ฆฌ, EDA
    • ODA ๊ทœ๋ชจ ํฌํ•จ ์—ฌ๋Ÿฌ ์ง€ํ‘œ๋ฅผ ๊ณ ๋ คํ•˜์—ฌ ์ •์ฑ… ์ฐธ์กฐ ๊ตญ๊ฐ€๋ฅผ ์„ ์ •, ๊ทธ๋“ค์˜ ODA ์ •์ฑ…์„ ๋น„๊ต๋ถ„์„
    • ๋‹จ๊ธฐ/์žฅ๊ธฐ ๊ด€์ ์—์„œ ํ•œ๊ตญ์˜ ODA ์ •์ฑ…์˜ ๋ฐฉํ–ฅ ์ œ์‹œ

Process

1. Feature Importance

  • Purpose: 42๊ฐœ ๊ตญ๊ฐ€์˜ 50์—ฌ ๊ฐœ ์ง€ํ‘œ(% of GDP)๋ฅผ ์ˆ˜์ง‘ํ•˜์—ฌ ODA ๊ทœ๋ชจ์— ์˜ํ–ฅ์„ ๋ฏธ์น˜๋Š” ์ฃผ์š” ๋ณ€์ˆ˜๋ฅผ ์ถ”์ถœ
  • Task: N๋…„์˜ Data๋กœ N+1๋…„์˜ ODA ๊ทœ๋ชจ๋ฅผ ์˜ˆ์ธกํ•˜๋Š” ํšŒ๊ท€ ๋ฌธ์ œ๋กœ ์ •์˜
  • Method: Filter Method, Embedded Method, Wrapper Method๋ฅผ ํ™œ์šฉํ•ด ์œ ์˜๋ฏธํ•œ ๋ณ€์ˆ˜ ์ถ”์ถœ, Baseline Model(Random Forest, Linear Regression)์˜ ์„ฑ๋Šฅ(MAE, MSE, $R^2$)์œผ๋กœ ๋ณ€์ˆ˜ ๊ฒ€์ฆ

2. Clustering

  • Purpose: ์ •๋Ÿ‰์ ์ธ ์ง€ํ‘œ๋ฅผ ํ™œ์šฉํ•ด ๊ตญ๊ฐ€๋ฅผ ๊ตฐ์ง‘ํ™”ํ•ด ์ •์ฑ…์„ ์ฐธ์กฐํ•  ๋งŒํ•œ ๊ตญ๊ฐ€๋ฅผ ์„ ์ •
  • Task: ๊ณ ์ฐจ์› ์ •๋Ÿ‰์  ์ง€ํ‘œ ๋ฐ˜์˜ (1) ODA ๊ทœ๋ชจ Data, (2) ODA ๊ด€๋ จ Data, (3) ์ฃผ์š” ๋ณ€์ˆ˜๊ฐ€ ํฌํ•จ๋œ Data ๊ด€์ ์—์„œ ๊ตญ๊ฐ€๋ฅผ ๋‚˜๋ˆ„๋Š” ๋ฌธ์ œ๋กœ ์ •์˜
  • Method: Hierarchical Clustering(Complete Link) ํ™œ์šฉ, Elbow Method๋กœ ๊ตฐ์ง‘ ๊ฐœ์ˆ˜ ์„ ์ •

Results

  • Feature Importance ๊ฒฐ๊ณผ 13๊ฐœ ์ฃผ์š” ๋ณ€์ˆ˜ ์„ ์ •, $R^2$ 0.9816 ๋‹ฌ์„ฑํ•˜์—ฌ ODA ๊ทœ๋ชจ๋ฅผ ๊ฒฐ์ •์ง“๋Š” ๊ตญ๊ฐ€ ์ง€ํ‘œ ํ™•์ •
  • Clustering ๊ฒฐ๊ณผ ์•„๋ž˜ 3๊ฐœ๊ตญ์„ Reference ๊ตญ๊ฐ€๋กœ ์„ ์ •
    • ์ผ๋ณธ: ๋†’์€ ์ ˆ๋Œ€์  ๊ทœ๋ชจ, ์ค‘๊ฐ„ ์ƒ๋Œ€์  ๊ทœ๋ชจ, ์ƒ์—…์ฃผ์˜์  ๋ชจ๋ธ, ํ•œ๊ตญ๊ณผ ๊ฐ™์€ ๋ฌธํ™”๊ถŒ
    • ๋ด๋งˆํฌ: ๋‚ฎ์€ ์ ˆ๋Œ€์  ๊ทœ๋ชจ, ๋†’์€ ์ƒ๋Œ€์  ๊ทœ๋ชจ, ์ธ๋„์ฃผ์˜์  ๋ชจ๋ธ, ์ด์ƒ์ ์ธ ODA ์ˆ˜ํ–‰
    • ์บ๋‚˜๋‹ค: ์ค‘๊ฐ„ ์ ˆ๋Œ€์  ๊ทœ๋ชจ, ์ค‘๊ฐ„ ์ƒ๋Œ€์  ๊ทœ๋ชจ, ํ•œ๊ตญ๊ณผ ๊ฐ™์€ Cluster
  • ํ•œ๊ตญ๊ณผ ์„ธ ๊ตญ๊ฐ€์˜ ์ •์ฑ…์„ ๋ถ„์„ํ•˜๊ณ  ๋น„๊ต(Report ์ฐธ๊ณ )

Conclusion

While Canada, Japan, and Denmark manage their ODA under a single agency, Koreaโ€™s MOFAT(Ministry of Foreign Affairs and Trade, ์™ธ๊ตํ†ต์ƒ๋ถ€) and MOSF(Ministry of Strategy and Finance, ๊ธฐํš์žฌ์ •๋ถ€) manage Grants and Loans, respectively. This leads to fragmentation and inefficiency, so a single agency should be in charge of ODA, with a single perspective from planning to evaluation. Denmark has the best ODA in terms of indicators, but it is difficult to accept due to the different public sentiment, culture, and geographical features of ODA with Korea. Therefore, it is necessary to take advantage of Canada and Japanโ€™s policy, which has the same culture as Korea, and make practical improvements, and ultimately aim for Denmarkโ€™s ODA. As a specific policy, we propose that Koreaโ€™s ODA should aim to ensure efficiency in aid distribution and management through the alignment of executive agencies, secure transparency through annual reports, and suggest an increase in the ODA/GNI ratio.

kdis-datascience-contest's People

Contributors

jwj51720 avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.