Aaron Baughman’s Post

Why should you consider taking part in Kaggle competitions? Once you do, how should you approach your projects to increase your chances of success? Geremie Yeo shares actionable insights based on his recent experiences in the Kaggle community.

🚀 I am excited to announce that I have completed the “Feature Engineering” certification from Kaggle! 🚀 This comprehensive course has provided me with valuable skills in feature engineering, a crucial aspect for enhancing the performance of my machine learning models. With this certification, I am now better equipped to achieve higher results in my online machine learning competitions on Kaggle. #MachineLearning #DataScience #FeatureEngineering #Kaggle #Certification #ContinuousLearning

Planning to head out to #SnowflakeSummit next month with the rest of the Satori team, and it's about to be wild. If you're there and want to learn about our product, and how it helps focus on AI & Analytics instead of access requests and security policies, send me a pm! Also, as an appetizer, here's a link to our raffle to win a Meta Quest 3! https://lnkd.in/ddzRNjNs

View my verified achievement from #Microsoft. Earners of this badge have completed the required steps of the Hackathon 2023 SkillUp #AI Challenge which include: registering for #Hackathon 2023 as a Hacker or Hacker/Advisor, and completed a 30-60 minute activity on the SkillUp AI Challenge tutorial website. The steps of the tutorial activity include creating a project and exporting a PowerPoint into a video file to upload to the project page.

Few things are more frustrating than building a product you're proud of, only to see your demo fall flat. Joseph George Lewis shows how you can avoid that fate with the aid of Gradio, a lightweight open-source package for making practical demos.

If you think large companies can't be agile, you're wrong. Here are three brilliant pieces of news from #Microsoft this week: 𝟭. 𝗕𝗿𝗶𝗻𝗴𝗶𝗻𝗴 𝘁𝗵𝗲 𝗳𝘂𝗹𝗹 𝗽𝗼𝘄𝗲𝗿 𝗼𝗳 𝗖𝗼𝗽𝗶𝗹𝗼𝘁 𝘁𝗼 𝗺𝗼𝗿𝗲 𝗽𝗲𝗼𝗽𝗹𝗲 𝗮𝗻𝗱 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀𝗲𝘀 We've expanded #Copilot for Microsoft 365 to businesses of all sizes and Introduced Copilot Pro. 𝟮. 𝗠𝗶𝗰𝗿𝗼𝘀𝗼𝗳𝘁 𝗔𝗜 𝗧𝗼𝘂𝗿 - 𝗕𝗲𝗿𝗹𝗶𝗻 𝗳𝗼𝗿 #𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿𝘀 | 𝗠𝗮𝗿𝗰𝗵 𝟭𝟵, 𝟮𝟬𝟮𝟰 Do you remember Steve Ballmer going crazy at a developer event? Even if I don't think we'll be quite that crazy, our AI Tour event will still be THE developer event in Germany this year. We will have >16 sessions (keynotes, breakouts, workshops), also some with our partner NVIDIA. And as a highlight, Eric Boyd, CVP AI Platform (he is the mastermind behind the partnership with OpenAI) will come to Berlin. 𝟯. 𝗘𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲 𝗣𝗹𝗮𝗻𝗲𝘁 𝗘𝗮𝗿𝘁𝗵 𝗜𝗜𝗜 𝘄𝗶𝘁𝗵 𝗠𝗶𝗻𝗲𝗰𝗿𝗮𝗳𝘁 You can now explore animals habitats by playing Minecraft. Go check out BBC Studios’ Planet Earth III.

View my verified achievement from Microsoft. Got to learn some more about AI through the SkillUp AI Challenge during Hackathon!

Alhamdulillah, another achievement unlocked is relevant to machine learning. Congratulations and thanks to my teammates. 🎉 🎊 Lastly, "Win or Lose no matter in any kind of competition all are experience and learning in our life." #datathon #machinelearning #reallifeproblem #competition #Robidatathon

A big thank you to everyone who joined our Machine Learning in R session! 🙌👩💻 - Engaging discussions on ML concepts 🗣️ - Hands-on experience with R demos 💻 - Valuable insights shared during the Q&A session 🤔🚀📊 Thank you for being part of this learning journey! 🌟👏 #DataScience #CommunityLearning #rstats

I'm thrilled to share that I've completed the Time Series course on Kaggle. This journey has equipped me with the skills to apply machine learning to real-world forecasting challenges using both parametric and non-parametric models. Throughout the course, I delved into the art of feature engineering, exploring time-step features, lag and lead features for serial dependence, moving average features, and trends that unveil the linear relationship between time and the time series. The course also sheds light on seasonal intricacies, introducing me to Seasonal Indicators, Fourier features, Cycles, and Autocorrelation. I even tackled engineering multioutput models for forecasting, employing techniques like linear regression, polynomial regression, and XGBoost. I also figured out how to make hybrid models. First, I use a simple model to capture the trend in the series, then fit a complicated model to the residuals to explain most of the remaining variance, which isn't captured by the first set of features and the simple model. Finally, I delved into Multistep Forecasting Strategies, exploring the Multioutput models, Direct strategy, and Recursive strategy—equipping me with the tools to forecast steps for real-time applications. It's been an exhilarating learning experience, and I'm eager to apply these newfound skills in the world of time series forecasting. #timeseries #timeseriesanalysis #dataanalysis #datascience #machinelearning #kaggle #dataanalytics