[{"authors":null,"categories":null,"content":"Florian Weiler is an associate professor in the Department of Public Policy at Central European University (Vienna), where he also directs the Applied Policy Project. His research centers on environmental issues — particularly climate adaptation and adaptation aid — and on the strategies stakeholders use to influence policy processes.\n","date":1782210600,"expirydate":-62135596800,"kind":"term","lang":"en","lastmod":1782210600,"objectID":"82bf1ec38621ab7405ec0d377789e60e","permalink":"https://pierrebeaucoral.github.io/authors/fweiler/","publishdate":"2026-06-01T00:00:00Z","relpermalink":"/authors/fweiler/","section":"authors","summary":"Florian Weiler is an associate professor in the Department of Public Policy at Central European University (Vienna), where he also directs the Applied Policy Project. His research centers on environmental issues — particularly climate adaptation and adaptation aid — and on the strategies stakeholders use to influence policy processes.\n","tags":null,"title":"Florian Weiler","type":"authors"},{"authors":null,"categories":null,"content":"I’m a development economist in training at CERDI. I spend most of my time debugging my R and Python codes trying to understand where “climate money” goes, what it is for, what it changes locally, and how data and ML can help answer these questions.\n","date":1782210600,"expirydate":-62135596800,"kind":"term","lang":"en","lastmod":1782210600,"objectID":"2525497d367e79493fd32b198b28f040","permalink":"https://pierrebeaucoral.github.io/authors/admin/","publishdate":"2026-06-01T00:00:00Z","relpermalink":"/authors/admin/","section":"authors","summary":"I’m a development economist in training at CERDI. I spend most of my time debugging my R and Python codes trying to understand where “climate money” goes, what it is for, what it changes locally, and how data and ML can help answer these questions.\n","tags":null,"title":"Pierre Beaucoral","type":"authors"},{"authors":null,"categories":null,"content":"Paul Vernus is a PhD candidate in development economics at CERDI (Université Clermont Auvergne) and an affiliate of FERDI.\nHis thesis explores international solidarity in the financing of climate-resilient development in Africa, looking in particular at how evolving partnerships and EU–Africa relations can support the most vulnerable countries and regions.\n","date":1773314100,"expirydate":-62135596800,"kind":"term","lang":"en","lastmod":1773314100,"objectID":"3a1903ee2eca3eb4e147b1e864a79aaf","permalink":"https://pierrebeaucoral.github.io/authors/pvernus/","publishdate":"2026-03-12T00:00:00Z","relpermalink":"/authors/pvernus/","section":"authors","summary":"Paul Vernus is a PhD candidate in development economics at CERDI (Université Clermont Auvergne) and an affiliate of FERDI.\nHis thesis explores international solidarity in the financing of climate-resilient development in Africa, looking in particular at how evolving partnerships and EU–Africa relations can support the most vulnerable countries and regions.\n","tags":null,"title":"Paul Vernus","type":"authors"},{"authors":null,"categories":null,"content":"Michaël Goujon is a professor of economics at Université Clermont Auvergne and a researcher at CERDI.\nHe works on development macroeconomics, with a particular focus on vulnerability, small island and overseas economies, and the links between climate change, resilience and development finance.\n","date":1735689600,"expirydate":-62135596800,"kind":"term","lang":"en","lastmod":1735689600,"objectID":"716e8e93187bd01b18690c8ac1a3f7f0","permalink":"https://pierrebeaucoral.github.io/authors/mgoujon/","publishdate":"2025-01-01T00:00:00Z","relpermalink":"/authors/mgoujon/","section":"authors","summary":"Michaël Goujon is a professor of economics at Université Clermont Auvergne and a researcher at CERDI.\nHe works on development macroeconomics, with a particular focus on vulnerability, small island and overseas economies, and the links between climate change, resilience and development finance.\n","tags":null,"title":"Michaël Goujon","type":"authors"},{"authors":null,"categories":null,"content":"Sébastien Marchand is an associate professor in economics at CERDI (Université Clermont Auvergne).\nHis research focuses on environmental and agricultural issues, climate and development, with empirical work in countries such as China, Viet Nam, Brazil and Cambodia.\n","date":1735689600,"expirydate":-62135596800,"kind":"term","lang":"en","lastmod":1735689600,"objectID":"d67b0a5181e59f8841a4728008816e0f","permalink":"https://pierrebeaucoral.github.io/authors/smarchand/","publishdate":"2025-01-01T00:00:00Z","relpermalink":"/authors/smarchand/","section":"authors","summary":"Sébastien Marchand is an associate professor in economics at CERDI (Université Clermont Auvergne).\nHis research focuses on environmental and agricultural issues, climate and development, with empirical work in countries such as China, Viet Nam, Brazil and Cambodia.\n","tags":null,"title":"Sébastien Marchand","type":"authors"},{"authors":["Pierre Beaucoral","Florian Weiler"],"categories":null,"content":"","date":1782210600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1782210600,"objectID":"ec76b8c5bf1175820810c663a6ffd99d","permalink":"https://pierrebeaucoral.github.io/talk/maladaptation-by-displacement-spatial-spillovers-of-adaptation-finance-in-sub-saharan-africa/","publishdate":"2026-06-01T00:00:00Z","relpermalink":"/talk/maladaptation-by-displacement-spatial-spillovers-of-adaptation-finance-in-sub-saharan-africa/","section":"event","summary":"Preliminary findings from joint work with Florian Weiler (CEU) on the spatial dimension of adaptation finance in Sub-Saharan Africa.","tags":["Adaptation finance","Maladaptation","Spatial spillovers","Sub-Saharan Africa"],"title":"Maladaptation by Displacement: Spatial Spillovers of Adaptation Finance in Sub-Saharan Africa","type":"event"},{"authors":null,"categories":null,"content":"This spring I spent a few weeks as a visiting PhD researcher at the Department of Public Policy and the Doctoral School of Political Science, Public Policy and International Relations at Central European University, on the Vienna campus.\nThe research stay I was hosted by Florian Weiler, whose work on the allocation of adaptation aid sits right at the border between economics and political science. That is also where much of my own PhD lives !\nDuring the visit I presented my work on climate aid, local economic activity and greenhouse-gas emissions in the DPP seminar, and started developing a new collaboration with Florian on adaptation aid in Africa.\nBeyond the seminar room, what stayed with me was how warmly the DPP and the doctoral school welcomed me into their community from day one. My sincere thanks to everyone there, and to the coordinators of the doctoral school for making the stay so easy and so productive.\nA deeper look on Central Europe… A research stay in Central Europe is also an invitation to explore it. On my days off I took the train out to Prague and Bratislava: two cities a stone’s throw from Vienna and impossible to resist.\nPrague gave me its castle and the spires of St. Vitus, the basilica at Vyšehrad, a grey-and-silver Vltava, and a sobering afternoon in the Museum of Communism (“dream, reality, nightmare”) which, for a development economist, is a useful reminder of what is at stake when institutions fail. Bratislava, smaller and quieter, offered pastel courtyards, a hilltop castle, and that particular pleasure of standing on a brass compass set into the pavement and reading off the distance to everywhere.\nHope to get back there some times !\n","date":1781049600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1781049600,"objectID":"673b69fd5707ef6562ba7fe655be7d91","permalink":"https://pierrebeaucoral.github.io/post/ceu-vienna/","publishdate":"2026-06-10T00:00:00Z","relpermalink":"/post/ceu-vienna/","section":"post","summary":"Notes from a spring research stay at the Department of Public Policy at Central European University, and exploring Prague and Bratislava on week-ends.","tags":["Research visit","Adaptation Finance","CEU","Travel"],"title":"A research visit to CEU Vienna - and a detour through Prague and Bratislava","type":"post"},{"authors":["Pierre Beaucoral","Florian Weiler"],"categories":null,"content":"","date":1780272e3,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1780272e3,"objectID":"a701fb057a52bf38f3b9b6a55076d52d","permalink":"https://pierrebeaucoral.github.io/project/maladaptation-spillovers/","publishdate":"2026-06-01T00:00:00Z","relpermalink":"/project/maladaptation-spillovers/","section":"project","summary":"Ongoing work with Florian Weiler (Central European University) on the spatial dimension of adaptation finance in Sub-Saharan Africa: how geocoded adaptation projects relate to outcomes in the areas around them. The project combines machine-classified adaptation finance with a place-based empirical design. Preliminary — full results to come; presented at the CERDI seminar in June 2026.\n","tags":["Ongoing work","Adaptation finance","Spatial spillovers","Sub-Saharan Africa","Causal inference"],"title":"Maladaptation by Displacement: Spatial Spillovers of Adaptation Finance in Sub-Saharan Africa","type":"project"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"","date":1777374e3,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1777374e3,"objectID":"32105c1ee3eebf93696112e94925de61","permalink":"https://pierrebeaucoral.github.io/talk/development-goals-emissions-costs-climate-and-development-finance-vis-a-vis-local-co-emissions/","publishdate":"2026-04-01T00:00:00Z","relpermalink":"/talk/development-goals-emissions-costs-climate-and-development-finance-vis-a-vis-local-co-emissions/","section":"event","summary":"Invited talk at the CEU Department of Public Policy seminar, presenting the emissions project during my visiting stay in Vienna.","tags":["Climate finance","Development finance","CO2 emissions","Causal inference","Machine learning"],"title":"Development Goals, Emissions Costs? Climate and Development Finance vis-à-vis Local CO₂ Emissions","type":"event"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"","date":1776556800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1776556800,"objectID":"df024af07e3e1ca7466df3071dee3467","permalink":"https://pierrebeaucoral.github.io/publication/cracking-the-code/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/publication/cracking-the-code/","section":"publication","summary":"Analysing development projects is crucial for understanding donors’ aid strategies, recipients’ priorities, and for assessing development finance capacity to address development issues through on-the-ground actions. In this area, the Organisation for Economic Co-operation and Development’s (OECD) Creditor Reporting System (CRS) dataset is a reference data source. This dataset provides a vast collection of project narratives from various sectors (approximately 5 million projects). While the OECD CRS provides a rich source of information on development strategies, it falls short in informing project purposes due to its reporting process, which is based on donors’ self-declared main objectives and predefined industrial sectors. This research aims to employ novel and reproducible approach for practitioners and researchers in social sciences that combines machine learning (ML) techniques, specifically natural language processing (NLP), with an innovative Python topic modelling technique called BERTopic, to categorise (cluster) and label development projects based on their narrative descriptions. By revealing existing yet hidden topics within development finance, this application of artificial intelligence enables a better understanding of donor priorities and overall development funding, and provides methods to analyse public and private project narratives.","tags":["Machine Learning","Development Finance"],"title":"Cracking the code: enhancing development finance understanding with artificial intelligence","type":"publication"},{"authors":null,"categories":null,"content":"EasyViz is a lightweight alternative to Streamlit-based dashboards. I built it as plain HTML + vanilla JavaScript + Chart.js, served statically; the browser talks to api.worldbank.org directly (permissive CORS), so no backend is required.\nFeatures Search — 100+ curated WDI indicators with fuzzy search via Fuse.js; optional one-click extension to the full ~20k-code taxonomy, cached locally for a week. Chart — line, area, or bar, with country presets (OECD, EU27, SSA, BRICS, G7, LDC, LMIC, G20, Default-20, All), adjustable year range, log scale, summary statistics, CSV export, and shareable URLs. Compare — scatter, bubble (size = third indicator), OLS regression with R², and Gapminder-style animated bubble chart. Upload — drop in a CSV or Excel file; the app auto-detects columns, reshapes wide to long, and resolves ISO3 codes. URL state — every control round-trips through query parameters, so any view is a permanent link. Embed mode — append ?embed=1 to strip chrome and drop the app into an iframe. Dark mode — respects prefers-color-scheme; toggleable. Keyboard — / focuses the search box. An optional FastAPI proxy (server-side cache, UCDP-authenticated routes) ships alongside for developers who want it.\n","date":1776470400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1776470400,"objectID":"5d44eab941564d3e454b3e2aeb97d2f9","permalink":"https://pierrebeaucoral.github.io/project/easyviz-static/","publishdate":"2026-04-18T00:00:00Z","relpermalink":"/project/easyviz-static/","section":"project","summary":"I developed a static web app to explore World Bank WDI indicators directly from the browser. Loads in under a second, no backend, no cold start. Search 100+ curated indicators (or the full ~20k taxonomy), chart line/area/bar with country presets (OECD, EU27, SSA, BRICS, G7, LDC, LMIC, G20), compare indicators (scatter, bubble, OLS with R², Gapminder-style animation), upload your own CSV/XLSX, and share every view through a URL.\n","tags":["Data visualization","Development economics","Open data","JavaScript","World Bank"],"title":"EasyViz — a static WDI explorer","type":"project"},{"authors":null,"categories":null,"content":"About this post Recently I have seen lot of coder, researcher, and other that were trying to find other solutions to benefit from AI than suscribing to a cloud AI/LLM. I have made some research, and I will talk about it here. This knowledge is quite volatile and might be outdated next month! However, if it mights help !\nWho this is for You want to use an AI assistant but you do not want to pay a monthly subscription, and you do not want your conversations, documents, or code sent to a company’s servers. This guide is for you.\nIt is not for people who need maximum performance at any cost. Local models are genuinely good now — but they are not GPT-4o or Claude Opus. I will be straight about that.\nThe summary upfront Running a large language model locally is now accessible to most people with a modern computer. The tooling has matured enormously in 2025–2026. If you have a recent Mac with an M-series chip, or a PC with a decent GPU, you can have a capable AI assistant running in under 10 minutes, completely offline, for free.\nThe catch: the experience is meaningfully worse than frontier cloud models, and the hardware requirements for the best local models are not trivial.\nHow it actually works A large language model is a file — a .gguf or similar format, typically 2–25 GB depending on the model. Your computer loads it into RAM or GPU memory and runs inference locally. No internet required after the initial download.\nThe tool that made this accessible to non-engineers is Ollama. It handles model downloading, memory management, and exposes a local API. You do not need to compile anything.\nGetting started: step by step 1. Install Ollama macOS / Linux:\ncurl -fsSL https://ollama.com/install.sh | sh Windows: download the installer from ollama.com.\nThat is it. Ollama runs as a background service.\n2. Pull a model ollama pull gemma3:4b # lightweight, fast, good for most tasks ollama pull gemma3:27b # much better quality, needs ~20 GB RAM ollama pull llama3.2:3b # Meta\u0026#39;s small model, very fast ollama pull mistral:7b # strong reasoning, good default ollama pull phi4-mini # excellent for coding, tiny footprint If not used to it, you have to run it in your terminal here is a nice Tutorial on how to use it\nStart with gemma3:4b or llama3.2:3b if you are unsure about your hardware.\n3. Chat immediately in the terminal ollama run gemma3:4b You are now talking to a local AI. No account. No API key. No data leaving your machine.\n4. Use a proper interface (optional but recommended) The terminal works, but a chat UI is more comfortable for daily use.\nOpen WebUI — the most polished option. Runs in your browser, connects to Ollama automatically.\ndocker run -d -p 3000:80 \\ -v open-webui:/app/backend/data \\ -e OLLAMA_BASE_URL=http://host.docker.internal:11434 \\ ghcr.io/open-webui/open-webui:main Then open http://localhost:3000.\nAlternatives: Chatbox, LM Studio (includes its own model downloader, no Docker required).\nHardware: the honest picture This is where most guides get vague. Here is the reality.\nRAM is the bottleneck Models run in RAM (or VRAM if you have a GPU). If the model does not fit in RAM, it spills to disk and becomes unusably slow.\nModel size RAM needed What you get 3B parameters ~3 GB Fast, limited reasoning, good for simple tasks 7–8B parameters ~6–8 GB Solid daily driver, handles most tasks 14–27B parameters ~12–20 GB Close to early GPT-4 quality 70B+ parameters ~48 GB+ Frontier quality — requires a workstation Practical baseline: 16 GB RAM gets you a good 7B model with room to breathe. 32 GB RAM gives you access to 27B models, which are genuinely impressive.\nApple Silicon is exceptional for this M1/M2/M3/M4 Macs have unified memory, meaning the CPU and GPU share RAM. A MacBook Pro M3 with 36 GB RAM runs a 27B model smoothly. This is not possible on most PC laptops. If privacy and local AI matter to you and you are buying hardware, Apple Silicon is currently the best value for this use case.\nNVIDIA GPUs on Windows/Linux If you have an NVIDIA GPU with enough VRAM (8 GB+), Ollama uses it automatically. A 4090 with 24 GB VRAM can run a 27B model at high speed. Older cards with less VRAM will offload layers to RAM, which works but is slower.\nNo GPU? It still works CPU-only inference is slow but usable for 3B–7B models. Expect 5–15 tokens per second on a modern CPU. Enough for writing and Q\u0026amp;A. Not great for real-time coding assistance.\nWhich models to actually use As of April 2026, these are the best options for local use:\nFor general use:\ngemma3:27b (Google, Apache 2.0) — excellent instruction following, long context (128K tokens), multilingual llama3.3:70b (Meta, Llama license) — best open-weight general model if you have the RAM mistral-small:22b — strong reasoning, good coding For coding:\nqwen2.5-coder:14b (Alibaba) — best dedicated coding model in this size range phi4:14b (Microsoft) — punches above its weight for reasoning and code For low-end hardware (≤8 GB RAM):\ngemma3:4b — best small model overall llama3.2:3b — …","date":177552e4,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":177552e4,"objectID":"602e1053990916274add180dd0d49fc3","permalink":"https://pierrebeaucoral.github.io/post/localllm/","publishdate":"2026-04-07T00:00:00Z","relpermalink":"/post/localllm/","section":"post","summary":"An Honest Guide for Privacy-Conscious Users","tags":null,"title":"Running LLMs Locally","type":"post"},{"authors":["Pierre Beaucoral","Paul Vernus"],"categories":null,"content":"","date":1773314100,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1773314100,"objectID":"add06159f4e2bc84c4ca4a36de658744","permalink":"https://pierrebeaucoral.github.io/talk/scaling-down-climate-adaptation-finance-justice-how-cosmopolitan-allocation-ideals-withstand-sovereign-realities-at-the-subnational-level/","publishdate":"2026-03-01T00:00:00Z","relpermalink":"/talk/scaling-down-climate-adaptation-finance-justice-how-cosmopolitan-allocation-ideals-withstand-sovereign-realities-at-the-subnational-level/","section":"event","summary":"Joint work with Paul Vernus on how global climate-justice ideals hold up when adaptation finance is allocated at the subnational level.","tags":["Adaptation finance","Climate justice","Subnational allocation","Geocoded data"],"title":"Scaling down climate adaptation finance justice: how cosmopolitan allocation ideals withstand sovereign realities at the subnational level","type":"event"},{"authors":["Pierre Beaucoral","Paul Vernus"],"categories":null,"content":"","date":1773273600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1773273600,"objectID":"33879e671ec4f5a2b65d9d29f9dd937a","permalink":"https://pierrebeaucoral.github.io/project/godad-adaptation-justice/","publishdate":"2026-03-12T00:00:00Z","relpermalink":"/project/godad-adaptation-justice/","section":"project","summary":"Joint work with Paul Vernus on how the justice ideals embedded in global climate governance play out when adaptation finance is allocated within countries. Using geocoded adaptation finance across tens of thousands of subnational units worldwide (2000–2022), the project contrasts cosmopolitan allocation principles with the sovereign realities that shape where funding actually lands. Work in progress — presented at the CERDI doctoral seminar (March 2026).\n","tags":["Ongoing work","Adaptation finance","Climate justice","Subnational allocation","Geocoded data"],"title":"Scaling down climate adaptation finance justice: how cosmopolitan allocation ideals withstand sovereign realities at the subnational level","type":"project"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"Bienvenue Ce site accompagne le cours Introduction à la méthode des doubles différences dispensé en master 2 – École d’Économie (Université Clermont Auvergne).\n","date":1768003200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1768003200,"objectID":"d108d1b847904bf49a8842580e1ae036","permalink":"https://pierrebeaucoral.github.io/course/gpe-did/","publishdate":"2026-01-10T00:00:00Z","relpermalink":"/course/gpe-did/","section":"course","summary":"Master-level course on modern DiD methods (French).","tags":["Teaching","Causal inference"],"title":"Introduction to Difference-in-Differences","type":"course"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"","date":1765463400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1765463400,"objectID":"2c9fa22bef3dc220ce62f431474df831","permalink":"https://pierrebeaucoral.github.io/talk/ce-que-lia-change-promesses-faits-doutes/","publishdate":"2025-01-01T00:00:00Z","relpermalink":"/talk/ce-que-lia-change-promesses-faits-doutes/","section":"event","summary":"Une analyse critique des transformations apportées par l'intelligence artificielle dans la recherche en sciences humaines : opportunités réelles, effets de mode et questionnements éthiques.","tags":["AI","Sciences Humaines","Méthodologie","Data"],"title":"Ce que l'IA change : Promesses, faits, doutes","type":"event"},{"authors":null,"categories":null,"content":"","date":1765411200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1765411200,"objectID":"12fa9de1809f161d07048854e3d64aee","permalink":"https://pierrebeaucoral.github.io/project/emissions/","publishdate":"2025-12-11T00:00:00Z","relpermalink":"/project/emissions/","section":"project","summary":"The global challenge of climate change is increasingly a place-based development challenge: investments that raise living standards also shape long-run emissions through energy systems, infrastructure, and urban form. This paper uses climate and non-climate development finance as a global laboratory to test whether development gains can be decoupled from CO2 emissions. I compile a global ADM2-level panel (2000–2022) linking geocoded aid project portfolios to local CO₂ emissions (EDGAR) and night-time lights (VIIRS). Climate-oriented portfolios are associated with small average declines in emissions—stronger where cumulative climate finance is large—while non-climate development finance systematically increases local emissions in the medium run. Night-time lights suggest expanding economic activity in mid-brightness regions, with muted effects in already bright places even as emissions rise. The results imply that meeting climate goals will require not only scaling climate finance but also greening mainstream development spending.\n","tags":["Ongoing work","Climate finance","Development finance","CO2 emissions","Causal inference","Machine learning"],"title":"Development Goals, Emissions Costs? Climate and Development Finance vis-à-vis Local CO₂ Emissions","type":"project"},{"authors":null,"categories":null,"content":" ✍️ Beyond Overleaf: Modern Alternatives for LaTeX Writing and Collaboration (2025 Update) By Pierre Beaucoral • October 2025\nFor many academics, Overleaf has long been the default platform for writing papers, reports, theses, and lecture notes in LaTeX. Its collaborative editing and cloud compilation were game-changers for research teams and classrooms.\nBut in late 2025, Overleaf introduced a significant limitation for free-tier users:\n⏱ A maximum compilation time of ~10 seconds per build.\nFor lightweight documents, this may be acceptable. But for most real academic workflows, theses, TikZ-heavy figures, large bibliographies, or multi-chapter manuscripts, 10 seconds is not enough. This has spurred a renewed interest in open-source, local, and alternative cloud-based solutions for LaTeX and scientific writing.\nThis post reviews and categorizes some of the best alternatives in 2025, grouped by workflow type:\n🧰 Local solutions for full control 🔁 Local + repository (hybrid) for collaboration and reproducibility 🌐 Online solutions for real-time collaboration (Overleaf-style) 🧰 Local solutions — Full control, no limits For researchers who prefer to work locally for speed, privacy, and reproducibility, these tools give you complete control over your typesetting environment.\n🖥️ VS Code + LaTeX Workshop What it is: Visual Studio Code + LaTeX Workshop Why it’s good: Full local compilation — no time limits. Continuous preview, spellcheck, snippets, and custom build recipes. Excellent integration with Git and BibTeX. Ideal for: solo researchers, or as a base for hybrid workflows with Git. ✍️ LyX LyX provides a structured, semi-WYSIWYG interface built on top of LaTeX. Great for theses or structured documents where you want to focus on content rather than code. Works offline and can sync via shared drives or Git for collaboration. 📄 R Markdown (.Rmd) and Quarto (.qmd) R Markdown and Quarto let you write academic manuscripts, reports, and books using Markdown syntax with embedded code (R, Python, Julia, etc.) and LaTeX under the hood. Output formats include PDF (via LaTeX), HTML, Word, and reveal.js slides. Quarto in particular is becoming a powerful alternative for scientific writing, enabling: Full reproducibility (code + text). Local compilation with no time limits. Version control through Git. Minimal Quarto example:\n--- title: \u0026#34;My Paper\u0026#34; author: \u0026#34;Pierre Beaucoral\u0026#34; format: pdf --- # Introduction This is written in Markdown but compiled through LaTeX. ✅ Ideal for academics who already use R or Python and want reproducible documents.\n✨ Typst — A modern, fast alternative to LaTeX Typst is a new, markup-based typesetting system designed to be as expressive as LaTeX with a cleaner, more consistent syntax. Compiles almost instantly, with a clean language that blends Markdown-like text and layout logic. Excellent documentation for LaTeX users migrating to Typst. Works entirely locally through the CLI, making it a strong candidate for researchers seeking better performance without abandoning the idea of typesetting. Tiny Typst snippet:\n#let title = \u0026#34;My First Paper\u0026#34; = title Hello *world*! Here\u0026#39;s a math equation: $E = mc^2$. ⚡ Key advantage: Typst documents typically compile in milliseconds, even for large projects.\n🔁 Local + repository (hybrid) — Collaborative and reproducible These workflows keep writing local, but use version control (GitHub/GitLab) for collaboration, history, CI builds, and archived artifacts.\n🧪 GitHub or GitLab + LaTeX pipelines Store .tex files in a repository. Set up Continuous Integration (CI) to compile on each push or pull request. Collaborators write locally but review changes and PDFs through the repo interface. Example GitHub Action for LaTeX:\nname: Build LaTeX on: [push, pull_request] jobs: build: runs-on: ubuntu-latest steps: - uses: actions/checkout@v3 - name: Install TeX Live run: sudo apt-get update \u0026amp;\u0026amp; sudo apt-get install -y texlive-full - name: Compile run: latexmk -pdf -interaction=nonstopmode main.tex ✅ Benefits:\nPotentially longer build times than browser editors (subject to CI provider limits). Full history and collaborative editing via pull requests. Ideal for papers with multiple co-authors comfortable with Git. 🌿 Quarto projects + Git repos For teams using Quarto (.qmd), putting your project in a GitHub or GitLab repo allows:\nEach author to compile locally. CI pipelines to generate the final PDF/HTML output. Easy versioning, code review, and reproducibility. --- title: \u0026#34;ggplot2 demo\u0026#34; author: \u0026#34;Norah Jones\u0026#34; date: \u0026#34;5/22/2021\u0026#34; format: html: fig-width: 8 fig-height: 4 code-fold: true --- Air Quality @fig-airquality further explores the impact of temperature on ozone level.\n#| label: fig-airquality #| fig-cap: \u0026#34;Temperature and ozone level.\u0026#34; #| warning: false library(ggplot2) ggplot(airquality, aes(Temp, Ozone)) + geom_point() + geom_smooth(method = \u0026#34;loess\u0026#34;) ⚡ Typst + Git + CI Typst can be built locally and compiled in CI pipelines with a few lines of YAML. Extremely fast builds make this …","date":1760572800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1760572800,"objectID":"7f1489396d8b5eec77414e40f5fd8b5e","permalink":"https://pierrebeaucoral.github.io/post/altexnatives/","publishdate":"2025-10-16T00:00:00Z","relpermalink":"/post/altexnatives/","section":"post","summary":"Modern Alternatives for LaTeX Writing and Collaboration","tags":null,"title":"Are there any AlTEXnatives?","type":"post"},{"authors":null,"categories":null,"content":"Bienvenue Ce site accompagne le cours Introduction à l’économétrie appliquée dispensé en Licence 3 – École d’Économie (Université Clermont Auvergne).\nCe cours de 30 heures (cours magistraux + travaux dirigés) propose une initiation pratique à l’économétrie, c’est-à-dire l’ensemble des méthodes statistiques utilisées pour tester des théories économiques ou analyser des données réelles.\nObjectifs À l’issue du cours, vous serez capable de :\nEstimer un modèle de régression linéaire (MCO) et interpréter les résultats. Tester les hypothèses classiques : normalité (Jarque–Bera), homoscédasticité (Breusch–Pagan, White, Goldfeld–Quandt), autocorrélation (Durbin–Watson, Breusch–Godfrey), exogénéité (Hausman, principe Nakamura \u0026amp; Nakamura). Corriger les problèmes détectés :\n– Moindres Carrés Généralisés (MCG),\n– corrections de White (hétéroscédasticité),\n– Cochrane–Orcutt (autocorrélation). Mettre en œuvre des variables instrumentales (2SLS) et tester la sur-identification (Sargan). Introduire des variables muettes ou des termes polynomiaux pour capter des effets non linéaires. Ressources du site Syllabus complet Notes de cours, TD et corrigés disponibles après chaque séance. Jeux de données et scripts EViews fournis dans l’ENT. Références principales Araujo, Brun \u0026amp; Combes (2008), Économétrie, Bréal — chapitres 1–2. Wooldridge (2009), Introductory Econometrics. Greene (2008), Econometric Analysis. Suivi du cours Prérequis : bases de statistiques (estimateurs, intervalles de confiance). Organisation : cours magistraux et TD en salle informatique (logiciel EViews). Évaluation : exercices de TD, mini-projets et examen final. Ce site réunit l’ensemble des documents utiles : syllabus, supports de TD, corrigés, jeux de données et bibliographie.\n","date":1757894400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1757894400,"objectID":"46d9cb9290b3ede4bd6b5c475ee50562","permalink":"https://pierrebeaucoral.github.io/course/econometrics/","publishdate":"2025-09-15T00:00:00Z","relpermalink":"/course/econometrics/","section":"course","summary":"Cours de 30 heures (L3 École d’Économie – UCA) présentant les bases de l’économétrie : modèle linéaire, tests d’hypothèses, corrections et variables instrumentales.","tags":null,"title":"Introduction à l’économétrie appliquée","type":"course"},{"authors":null,"categories":null,"content":"Bienvenue Ce site accompagne le cours Modélisation statistique dispensé en master 2 – École d’Économie (Université Clermont Auvergne).\n","date":1757894400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1757894400,"objectID":"e8d78838876fa96a24f54324bb73d903","permalink":"https://pierrebeaucoral.github.io/course/modelisation-statistique/","publishdate":"2025-09-15T00:00:00Z","relpermalink":"/course/modelisation-statistique/","section":"course","summary":"Cours à destination des M2 Développement durable et économie de la santé","tags":null,"title":"Modélisation Statistique","type":"course"},{"authors":null,"categories":null,"content":" Why this topic? As a French researcher, I was particularly struck by the wildfires in Aude this year, as well as those affecting Spain and Portugal. These events prompted me to take a closer look at the geography of fire activity in Europe. My objective was to get a clearer picture of where fires are occurring, when they peak, and how 2025 compares with previous exceptional fire seasons.\nWhat does the data show? Using EFFIS rapid mapping perimeters (fires ≥ 30–50 ha), the analysis reveals that burn scars are overwhelmingly concentrated in Southern and Eastern Europe, with Spain, Portugal, Italy, and Greece among the most severely affected. France also shows a marked increase compared to previous years, confirming that wildfire risk is extending northward along the Mediterranean arc.\nIn terms of timing, 2025 fires followed a late-summer pattern. Activity was moderate in June, rose sharply in July, and culminated in August with the largest burned areas, consistent with the idea that prolonged heat and depleted water reserves create peak vulnerability toward the end of the season. Daily counts also reveal three distinct waves of fire activity, in early July, late July, and early August, likely corresponding to short-term climatic triggers such as heatwaves and wind events layered on cumulative dryness.\nHow does 2025 compare? To put 2025 in context, I compared it with other major fire years. 2017 remains an important benchmark: Portugal recorded catastrophic losses that year, while Spain also suffered heavily. By contrast, 2025 is dominated by Spain, which surpasses its previous records, while Portugal’s losses are smaller but still substantial. Adding 2022 and 2023 highlights that extreme fire years now recur more frequently, each with its own epicenter, Greece in 2022–2023, Spain in 2025, Portugal in 2017, showing how exposure rotates across the Mediterranean and Balkans under shifting climatic and local conditions.\nLimitations These results are based on EFFIS rapid mapping data and should be read as estimates, not exact totals. National fire statistics may differ due to thresholds, seasonal definitions, or inclusion of smaller fires. Here, “summer” is defined as June–August, which excludes late-season events (e.g., in September), and country attribution is based on maximum overlap, which can misrepresent border regions. The purpose of this work is analytical and exploratory; for operational decisions, official national and European statistics remain the reference.\nThe blog is available Here\n","date":1757376e3,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1757376e3,"objectID":"fbb5bfcf5b6cc406c10f17c44564a264","permalink":"https://pierrebeaucoral.github.io/post/fire/","publishdate":"2025-09-09T00:00:00Z","relpermalink":"/post/fire/","section":"post","summary":"Triggered by the devastating fires in Southern France and the Iberian Peninsula, I explored recent EFFIS data to better understand the geography and timing of wildfires across Europe.","tags":null,"title":"Wildfire dynamics in Europe, Summer 2025","type":"post"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"Slides available upon requests.\n","date":1751626800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1751626800,"objectID":"6c03a5e19ec2f53f9ae6a63ea7572aa5","permalink":"https://pierrebeaucoral.github.io/talk/icde-2025/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/talk/icde-2025/","section":"event","summary":"No plan, No Aid? The effects of National Adaptation Plan implementation on received Adaptation Aid","tags":["Ongoing work","Machine learning","Development finance","Adaptation finance"],"title":"ICDE 2025","type":"event"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"Slides available upon requests.\n","date":1739455200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1739455200,"objectID":"3514730fe5da70d97ce5bf22f0b93965","permalink":"https://pierrebeaucoral.github.io/talk/miners-meeting/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/talk/miners-meeting/","section":"event","summary":"Leveraging Artificial Intelligence for Enhanced Insights into Development Finance: A Focus on Climate Finance estimation","tags":["Ongoing work","Machine learning","Development finance","Climate Finance"],"title":"Miners meeting","type":"event"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"Slides available upon requests.\n","date":1738072800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1738072800,"objectID":"a3e1a87d8d7d128435a81ce066e51b8e","permalink":"https://pierrebeaucoral.github.io/talk/afedev-jdd/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/talk/afedev-jdd/","section":"event","summary":"Under the Green Canopy: bringing up to date public climate finance determinants analysis with AI","tags":["Ongoing work","Machine learning","Development finance","Climate Finance"],"title":"AFEDEV JDD","type":"event"},{"authors":["Pierre Beaucoral","Michaël Goujon","Sébastien Marchand"],"categories":null,"content":"","date":1735689600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1735689600,"objectID":"7178c944410b8c18bd31057c4258b867","permalink":"https://pierrebeaucoral.github.io/project/nap/","publishdate":"2025-01-01T00:00:00Z","relpermalink":"/project/nap/","section":"project","summary":"“No Plan, No Aid?” examines whether implementing a National Adaptation Plan (NAP) increases a country’s climate adaptation finance. Through a novel theoretical model and a robust Double Machine Learning (DML) approach, the authors find that NAP adoption does send mixed signals—reducing perceived vulnerability while boosting perceived capacity. Crucially, they find that while traditional econometric models show no impact or even a negative one, modern methods reveal a significant positive effect on adaptation aid.\n","tags":["Ongoing work","Development finance","Climate Finance"],"title":"No plan, No Aid? The effects of National Adaptation Plan implementation on received Adaptation Aid","type":"project"},{"authors":null,"categories":null,"content":" What is the Economic Policy Management (GPE) training programme? The Economic Policy Management (GPE) training programme is a leading French course for developing public economic capacity in stakeholders from developing countries, with a particular focus on French-speaking countries in sub-Saharan Africa. It leads to a Master 2, a national diploma delivered by the University of Clermont Auvergne (UCA).\nWhat was the course about? The objective of this course was to equip them with the skills to utilise R programming for data analysis and visualisation, including basic statistics and visualisation. The final course comprised an exercise in which they selected a key data set related to their field of expertise to highlight. In only 18 hours discovering this tool, they manage to get some remarquable progress! here are some of the results:\nGraphs by Bio Bertrand Mama Akouvi Sophie Agbavo and Biri Salatou Diagana ","date":1728604800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1728604800,"objectID":"f7b4254ab62096f995bdff2f50bd1fce","permalink":"https://pierrebeaucoral.github.io/post/gpe/","publishdate":"2024-10-11T00:00:00Z","relpermalink":"/post/gpe/","section":"post","summary":"I was delighted to have the opportunity to work with the members of the Economic Policy Management (GPE) training programme, providing instruction in the use of R programming.","tags":null,"title":"Teachings with the Economic Policy Management (GPE) training programme","type":"post"},{"authors":null,"categories":null,"content":" Course outline This course is designed for those with no prior experience of data analysis. This 18-hour course on R covers loading datasets, performing basic statistics, and creating data visualizations. The course is divided into 6 3-hour sessions.\nRequired files Data required can be downloaded here, course with code and explanation can be downloaded here.\nAlso, few slides on how to make visualization in the case of GPE student.\n","date":1725926400,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1725926400,"objectID":"2c58f99769caf740c2eebeadf784178c","permalink":"https://pierrebeaucoral.github.io/course/r-for-beginners/","publishdate":"2024-09-10T00:00:00Z","relpermalink":"/course/r-for-beginners/","section":"course","summary":"This course is designed for those with no prior experience of data analysis. This 18-hour course on R covers loading datasets, performing basic statistics, and creating data visualizations. The course is divided into 6 3-hour sessions.","tags":null,"title":"R programming for beginners","type":"course"},{"authors":null,"categories":null,"content":"Further Explanations Climate finance plays a crucial role in combating climate change, addressing mitigation, adaptation, and environmental goals. However, traditional methods of analyzing and categorizing these financial flows often rely on self-reported data and broad categorizations, which can lead to inaccuracies and overreporting.\nThis study uses ClimateFinanceBERT, an advanced AI-based approach to analyze and classify climate-related financial flows with unprecedented precision. Leveraging machine learning techniques, the study processes over 1.3 million development finance projects from the OECD Creditor Reporting System (CRS), identifying climate-relevant projects and categorizing them into mitigation, adaptation, and environmental domains. The two-stage classification approach significantly enhances the accuracy of estimating climate finance and its determinants compared to traditional methods like the Rio markers.\nKey Findings Adaptation Finance:\nAllocation is increasingly driven by recipient vulnerability and institutional capacity, prioritizing small island developing states (SIDS) and least-developed countries (LDCs). Historical ties, such as colonial relationships or shared languages, show little to no significance in influencing adaptation finance. Countries with strong governance and public administration frameworks receive more adaptation aid, highlighting the importance of institutional capacity over fiscal management.\nMitigation Finance:\nWhile historical ties like colonial relationships and shared languages still play a role, their influence has significantly weakened under more rigorous AI-driven analysis. Mitigation funding decisions align more closely with quantifiable outcomes, such as greenhouse gas reductions, resulting in a standardized allocation process. Governance indicators are less critical compared to adaptation finance, reflecting the technical and infrastructure-focused nature of mitigation projects.\nDisparities in Traditional Estimates:\nThe study reveals significant overreporting in conventional climate finance estimates, particularly for adaptation projects. This is attributed to methodological biases in traditional classification systems, which often categorize non-climate projects as climate-related. Using ClimateFinanceBERT, the study identifies a ten-year delay in achieving the $100 billion annual global climate finance target based on adjusted estimates, with the target potentially reached only by 2032 under current trends.\nImplications for Policy and Practice The findings underline the need for more nuanced and data-driven approaches to climate finance allocation. Policymakers and stakeholders should consider the following:\nDifferentiated Strategies: Adaptation finance should prioritize vulnerability and capacity-building, while mitigation finance can benefit from standardized allocation frameworks aligned with measurable targets.\nImproved Transparency: The reliance on outdated classification systems like the Rio markers highlights the need for adopting AI tools like ClimateFinanceBERT to enhance the accuracy and accountability of climate finance reporting.\nCapacity Building: Donors should invest in strengthening institutional capacities in vulnerable regions to improve the effectiveness of adaptation projects.\nThis study demonstrates the transformative potential of AI in analyzing climate finance, paving the way for more equitable and impactful distribution of resources. Future research should expand this methodology to multilateral and private financial flows to provide a comprehensive view of climate finance dynamics and ensure resources align with global sustainability goals. Working Paper available upon request from January 2025.\n","date":1725580800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1725580800,"objectID":"c14edd4f2d8b70adb5adb5fdb64f44b3","permalink":"https://pierrebeaucoral.github.io/project/climate-finance-estimation/","publishdate":"2024-09-06T00:00:00Z","relpermalink":"/project/climate-finance-estimation/","section":"project","summary":"Climate finance is critical for addressing the multifaceted challenges of climate change, encompassing mitigation, adaptation, and environmental sustainability. This study aims to renew the analysis of a critical part of climate finance determinants' allocation across these dimensions and accurately estimate bilateral public climate finance flows using an advanced machine learning approach. ClimateFinanceBERT (Bidirectional Encoder Representations from Transformers) is employed to classify development finance projects, distinguishing those that contribute to climate mitigation, adaptation, and environmental objectives. By examining a comprehensive dataset of development finance projects (OECD CRS) and replicating a recent research on climate public aid determinants, this study identifies key factors influencing the allocation of climate finance. This work updates significant patterns in climate finance distribution. This research contributes to the growing field of climate finance by offering a robust analytical framework for assessing the determinants of climate finance and proposing a scalable solution for monitoring financial flows aimed at addressing climate change in its entirety. The insights gained have important implications for policymakers and stakeholders striving to understand and optimize the allocation of climate finance to support global sustainability and resilience goals.\n","tags":["Ongoing work","Machine learning","Development finance","Climate Finance"],"title":"Under the Green Canopy: bringing up to date public climate finance determinants analysis with AI","type":"project"},{"authors":null,"categories":null,"content":" An AI powered podcast Want to learn more about development finance and categorization of development projects? Here you can find a quick summary of my research project:\nhttps://github.com/user-attachments/assets/5e11695c-2b9d-40a2-85de-9b47e95f9f15\nAn AI powered rap song More informal or funnier way to get an idea of the same research project thanks to suno? here it is!\nWhat is Suno? Suno AI, or simply Suno, is an innovative artificial intelligence music creation program designed to produce lifelike songs incorporating vocals and instrumentation, or solely instrumental compositions. Widely accessible since December 20, 2023, Suno was introduced through a web application launch and a strategic partnership with Microsoft, which integrated Suno as a plugin in Microsoft Copilot. This program operates by generating songs based on textual prompts provided by users. While Suno keeps its training dataset undisclosed, it assures users of measures taken to prevent plagiarism and copyright infringements.\nHow did I use it? I use the summary of my research project on NLP and the OECD CRS dataset and give some instructions about the style (initially “old emotional rap”), then I let the machine do the maths and here is the result:\nhttps://github.com/PierreBeaucoral/PierreBeaucoral.github.io/assets/148867967/5545a591-bea3-4e60-ae43-fc6a27cab866\n","date":1717545600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1717545600,"objectID":"c34cc97acb956dc3271f68c61984d18f","permalink":"https://pierrebeaucoral.github.io/post/suno/","publishdate":"2024-06-05T00:00:00Z","relpermalink":"/post/suno/","section":"post","summary":"Or a little break using Suno to sum up a research project.","tags":null,"title":"Asking AI to vulgarise an AI-powered research project","type":"post"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"Slides available upon requests.\n","date":1715094e3,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1715094e3,"objectID":"798b61426119302b608b5abb3d2ee265","permalink":"https://pierrebeaucoral.github.io/talk/economia-conference/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/talk/economia-conference/","section":"event","summary":"Leveraging Natural Langage processing techniques for a better classification of development projects: a case study using OECD CRS dataset.","tags":["Ongoing work","Machine learning","Development finance"],"title":"Econom'IA conference","type":"event"},{"authors":null,"categories":null,"content":" The econom’ia conference “This conference aims to explore and foster the cutting-edge applications of Artificial Intelligence (AI), Text Mining, Web Mining, Data Visualization, and other innovative techniques in all the fields of Economics. Econom’IA brings together researchers from the academic world as well as entrepreneurs that use innovative techniques to analyse economic data.\nThis 2-day conference proposes training workshops in the morning to discover new tools and techniques. The afternoon is devoted to presentations and discussions of articles using at least one of the innovative techniques covered in the workshops.\nThe 2 days are led by leading researchers in the field.”\nMy own participation I had the opportunity to attend the whole conference, but I will talk more about the second day when I presented my own work!\nSecond day of the conference I recently had the pleasure of attending a fantastic training and presentation event. The day started with a warm welcome and greetings from 8.45am to 9am.\nFrom 9am to 12.30pm, we were treated to an excellent introductory course on machine learning led by Mathieu Bernard, a CNRS research engineer at Economix. The hands-on course covered the basic concepts of machine learning, including building predictive models, data preparation, learning and evaluation. It was a great opportunity to brush up on our Python skills.\nAfter a delicious lunch break from 12.30pm to 2pm, the afternoon was dedicated to presentations from esteemed speakers in the field.\nFirst, from 14:00 to 15:00, we heard from keynote speaker Emmanuel Flachaire from the University of Aix-Marseille on “Interpretability and causality in machine learning”. It was a fascinating and thought-provoking talk that managed to build bridges from machine learning to academic research in economics, especially for an econometric approach when presenting his last paper with co-authors GAM(L)A: An econometric model for interpretable Machine Learning.\nFrom 15:00 to 15:45 I had the honour to present my work on “Cracking the Code: Enhancing Development Project Classification with NLP on OECD CRS data”. It was a great opportunity to share my research and insights with the community and receive valuable feedback and questions.\nThere was a short break from 15:45 to 16:00 before Johannes van der Pol of the Copernicus Institute for Sustainable Development, Utrecht University, took the stage from 16:00 to 16:45 to discuss “One patent to rule them all? The scope of multi-standard essentials explored”. It was a very informative and engaging presentation that sparked interesting discussions and debates about the machine learning techniques used and the overall research question.\nFinally, from 16:45 to 17:30, Rim Bahroun from EconomiX, Université Paris Nanterre, presented on “Large Langage Models for advanced text processing: application in economics”. It was a great way to end the day with a fascinating presentation that summarised the progress of techniques in natural language processing.\nOverall, the training and presentation event was a great success and a fantastic opportunity to expand our knowledge of machine learning and its applications. I would highly recommend attending future events like this to anyone interested in the field and would like to thank the organising committee for this first successful edition.\n","date":171504e4,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":171504e4,"objectID":"2d35be7d1e5828cf8a824eb489b72544","permalink":"https://pierrebeaucoral.github.io/post/economia/","publishdate":"2024-05-07T00:00:00Z","relpermalink":"/post/economia/","section":"post","summary":"I was pleased to take part and to present one of my work at this conference!","tags":null,"title":"Participation to the Econom'IA conference","type":"post"},{"authors":null,"categories":null,"content":" A new centre for development and development activities This event marked the official launch of a new centre of excellence in Clermont-Ferrand, bringing together CERDI, Ferdi and the Global Development Network to develop solutions for developing countries and bring together academic researchers, policy makers and project implementers. I am proud to be part of this laboratory.\nA day full of meetings This event was full of insightful interventions, speeches and meetings. I had the opportunity to discuss their vision of development policy and climate action with many senior officials from the French Ministry of Finance, the French Development Agency and, of course, Chrysoula Zacharopoulou, Minister of State for Development and International Partnerships, attached to the Minister of European and Foreign Affairs.\nPhoto by Jonathan Sarago and Emilie Connois Photo by Jonathan Sarago and Emilie Connois\n","date":1712188800,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1712188800,"objectID":"6796004b615dd12b3d10b29a04859785","permalink":"https://pierrebeaucoral.github.io/post/pcdi/","publishdate":"2024-04-04T00:00:00Z","relpermalink":"/post/pcdi/","section":"post","summary":"I was pleased to take part of the event of the creation of the PCDI","tags":null,"title":"Creation of the Pôle Clermontois de Développement International","type":"post"},{"authors":null,"categories":null,"content":"","date":1711065600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1711065600,"objectID":"2fc08020c8dc1bbe623878069ad0a039","permalink":"https://pierrebeaucoral.github.io/project/crs-ml/","publishdate":"2024-03-22T00:00:00Z","relpermalink":"/project/crs-ml/","section":"project","summary":"","tags":["Ongoing work","Machine learning","Development finance"],"title":"Cracking the Code: Enhancing Development Project Classification with NLP on OECD CRS data.","type":"project"},{"authors":["Pierre Beaucoral"],"categories":null,"content":"Slides available upon requests\n","date":1703073600,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1703073600,"objectID":"873564d865f85bd7340e661e10b58481","permalink":"https://pierrebeaucoral.github.io/talk/cerdi-phd-seminar/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/talk/cerdi-phd-seminar/","section":"event","summary":"Better classification of international development finance for a better understanding of climate finance: the role of machine learning.","tags":["Ongoing work","Machine learning","Climate finance"],"title":"CERDI PhD seminar","type":"event"},{"authors":["Jean-Michel Severino","Pierre Beaucoral"],"categories":null,"content":"","date":1684195200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1684195200,"objectID":"baa61ce0d37eb496d671a12de8d6d2ef","permalink":"https://pierrebeaucoral.github.io/publication/dfi-private-sector-africa/","publishdate":"2017-01-01T00:00:00Z","relpermalink":"/publication/dfi-private-sector-africa/","section":"publication","summary":"Redacted appendix for this insightfull article from Jean-Michel Severino. \nThe article below argues for the need to strongly accelerate public involvement in support of entrepreneurial emergence in poor and fragile countries. After mentioning the economic and employment issue, it explains how this priority has long disappeared from the international agenda as well as from domestic public policies, particularly in Africa. Efforts to promote the private sector have in practice focused on foreign direct investment and the largest companies. Middle- and emerging income countries, and a limited number of sectors and financial instruments, such as debt, have been valued. The article evokes the gradual change of perception on this subject from the beginning of the century and the emergence of new so- called impact actors focused particularly on SMEs in poor countries, accompanied by some public private sector financing institutions (DFIs), development agencies or foundations.\n{style=\"text-align: justify;\"}","tags":["Development finance","Private sector"],"title":"Are development finance institutions supporting the private sector in Africa?","type":"publication"},{"authors":["Jean-Michel Severino"],"categories":null,"content":"","date":1461715200,"expirydate":-62135596800,"kind":"page","lang":"en","lastmod":1461715200,"objectID":"ce78a12200bca410316e75d08652705a","permalink":"https://pierrebeaucoral.github.io/project/billions-millions/","publishdate":"2016-04-27T00:00:00Z","relpermalink":"/project/billions-millions/","section":"project","summary":"Publication from Jean-Michel Severino (https://ferdi.fr/biographies/jean-michel-severino) where I provided an appendix aiming to reveal the actual support provided by development finance institutions to the African private sector.\nThe article below argues for the need to strongly accelerate public involvement in support of entrepreneurial emergence in poor and fragile countries. After mentioning the economic and employment issue, it explains how this priority has long disappeared from the international agenda as well as from domestic public policies, particularly in Africa. Efforts to promote the private sector have in practice focused on foreign direct investment and the largest companies. Middle- and emerging income countries, and a limited number of sectors and financial instruments, such as debt, have been valued. The article evokes the gradual change of perception on this subject from the beginning of the century and the emergence of new so- called impact actors focused particularly on SMEs in poor countries, accompanied by some public private sector financing institutions (DFIs), development agencies or foundations..","tags":["Private sector","Development finance"],"title":"Millions for billions: Accelerating African entrepreneurial emergence for accelerated, sustainable and job-rich growth","type":"project"}]