If you ever felt like this, or even if you haven’t but you know you are down the path that I was - updating software without a care in the world for your production code. The good news is I eventually learned Docker, Git, and a lot more. …when you are about to present an awesome Shiny app to 500 data scientists and the most powerful folks in finance. Learning tools like Docker & Git can save you countless hours when in production or… Disaster averted.īut this single problem that cost me a day of work AND taught me a valuable lesson. One of the args to a function changed names. I muster a “no thanks.” And try not to scream. Miles away from home.īeads of sweat instantly started to form.
#Que es el intervalo de confianza how to#
This was before I knew how to use GitHub & Docker like I do now. My shiny app suddenly stopped working and I went into straight panic-mode. I had a presentation for R/Finance where I was going to present a Shiny App that I was working on. What stresses me out isn’t the same thing that stressed my parents out.
It will save you years on your data science journey. PS - I'll leave a link to my free 40-minute webinar in the comments.
Your friendly neighborhood data scientist, I cover all of it in one value-packed training. Literally what worked, and what mistakes to avoid in your path. I put together a free webinar - It's 40-minutes of training on the 10 secrets that took my 5-years to learn in my path to becoming a data scientist. Like 5-years.Īnd, I don't want you to have to take EONS to learn too. It took me painfully long to learn secrets like this. You just have to know how to use these algorithms. These algorithms penalize bad features so we don't have to drop them cold-turkey style like with Linear Regression. But, to more accurate forecasts and churn models, I began using new algorithms like XGBoost & H2O.Īnd what's cool is these algorithms will help you. This magical p-value was actually costing me good information?!!! I started learning machine learning and actually seeing that my model performance would improve with features that had p-values greater than 0.05. So naturally I'd do whatever I could to get my p-values under 0.05. If p>0.05, that feature gets axed cold turkey. I always thought it was funny how in the stats books, they'd always tell you of this magical P-Value level of 0.05. This used to be me when I was first learning linear regression. Your Friendly Neighborhood Data Scientist | 3000+ students learning Data Science for Business | 65K+ Followers
0 kommentar(er)
