If playback doesn't begin shortly, try restarting your device.
•
You're signed out
Videos you watch may be added to the TV's watch history and influence TV recommendations. To avoid this, cancel and sign in to YouTube on your computer.
CancelConfirm
Share
An error occurred while retrieving sharing information. Please try again later.
Having data analysis and visualization skills is increasingly important in the new age of Large Language Models and generative AI. But how does a non-Python developer skill up rapidly with the tools & best practices required to achieve project goals, without having the benefit of years of Python or data science experience? This is where the right developer tooling, with a little bit of AI assistance, can help.
In this talk, we'll go from identifying an open-source data set, to analyzing it for insights and visualizing relevant outcomes, in 25 minutes - with just a GitHub account and an OpenAI endpoint. Along the way, we'll introduce you to a series of developer tools that make your journey easier:
…...more
Simplifying Data Analysis & Visualization with Developer Tools & AI | Python Data Science Day
2Likes
410Views
2024Mar 28
Having data analysis and visualization skills is increasingly important in the new age of Large Language Models and generative AI. But how does a non-Python developer skill up rapidly with the tools & best practices required to achieve project goals, without having the benefit of years of Python or data science experience? This is where the right developer tooling, with a little bit of AI assistance, can help.
In this talk, we'll go from identifying an open-source data set, to analyzing it for insights and visualizing relevant outcomes, in 25 minutes - with just a GitHub account and an OpenAI endpoint. Along the way, we'll introduce you to a series of developer tools that make your journey easier:
Open Dataset: to ""analyze"" - from Kaggle, Hugging Face, or Azure
Data Wrangler: to ""sanitize"" data - extension from Visual Studio Code
Jupyter Notebook: to ""record"" process - for transferable learning
GitHub Codespaces: to ""pre-build"" environment - for consistent reuse
GitHub Copilot: to ""explain/fix"" code - for focused learning with AI help
Microsoft LIDA: to ""suggest/build"" visualization goals - for building your intuition with AI help
The talk comes with an associated repo that you can fork - then replace with your own dataset to extend or experiment on your own later. By the end of the talk you should have a sense of how you can go from discovering a data set to getting some visual insights about it, using existing tools with a little AI assistance.
Chapters:
00:00 Simplifying Data Analysis & Visualization with Developer Tools & AI
00:29 Follow along
00:54 Introduction - Data Analysis Challenges & Goals
04:44 GitHub Codespaces - Reusable environments
08:32 Jupyter Notebooks - Make it reproducible
11:18 GitHub Copilot - AI-assisted learning
14:43 Visual Studio Code - Productivity extensions
15:39 Open Datasets - Data Wrangler
19:15 Resonsible AI toolkit - Model debugging for fairness
21:13 Project LIDA - AI-assisted intuition & visualization
25:24 Azure AI Studio - Paradigm shift to LLM Ops
25:47 Summary - Questions & Next Steps
Resources:
30 Days of Data Science: https://30daysof.github.io/data-scien...
Data Science Recipes: https://aka.ms/2024/data-science-recipes
Workshops: https://aka.ms/workshops/python-data-...
Survey https://aka.ms/Python/DataScienceDay/...
Python at Microsoft https://aka.ms/python
Cloud Skills Challenge - through April 15, 2024
https://aka.ms/Python/DataScienceDay/CSC
GitHub codespaces https://github.com/codespaces
VS Code Release notes https://code.visualstudio.com/updates
Featuring: Nitya Narasimhan, PhD, Senior AI Advocate, Microsoft (@nitya)…...more