Uncertainty in Statistical Modeling Explained Intuitively
HTML-код
- Опубликовано: 16 июн 2021
- One of the main goals of statistics is to help make predictions. That could be predictions about how effective a new drug is in stopping a disease, how likely a hurricane is to devastate your city, or how likely an investment is to be profitable. Intuitively, we probably get that all those predictions have some uncertainty around them: drugs don’t always work, hurricanes sometimes act unpredictably, and the stock market goes down for no apparent reason. When data scientists build models to make these types of predictions they fully appreciate that uncertainty is just a fact of life. They also appreciate that the other side of that coin is confidence in prediction accuracy.
Welcome to Data Demystified. I’m Jeff Galak and in this episode we’re going to learn to intuitively understand how predictive models incorporate uncertainty. We’ll also learn what role prediction confidence has and how it is related to, but not quite the same thing as uncertainty.
Link to video about forecasting hurricane paths: • Predicting Hurricane P...
Follow me at:
LinkedIn: / jeff-galak-768a193a
Patreon: / datademystified
Equipment Used for Filming:
Nikon D7100: amzn.to/320N1FZ
Softlight: amzn.to/2ZaXz3o
Yeti Microphone: amzn.to/2ZTXznB
iPad for Teleprompter: amzn.to/2ZSUkNh
Camtasia for Video Editing: amzn.to/2ZRPeAV
Thankyou very much sir❤
This reminds me of the "confidence interval" and "prediction interval". The Prof also gave us a quick intuition explanation: if you care about the range/prediction noise, check the prediction interval, but if you care about mean, check the confidence interval. My intuition is that for a given X, the range of possible Ys is the "prediction interval", and the range of the possible mean of Ys is the "confidence interval". What's your understanding of this? How should we differentiate the two concepts? Thank you! And Happy New Year!
Thank you so much for make it simple.
Could we have another video to explain the concept of entropy?!
Excellent explanation! I’m finally beginning to grasp the relevance and application of the confidence concept in stats. Thank you. Keep up the great teaching.
Great Explanation! Can you explain when someone says that there is a 95% chance that is going to rain tomorrow, how to interpret that?
Great question. I actually have a video covering that as well: ruclips.net/video/81aEUYr9mcA/видео.html
The short version is that when we're dealing with single events like tomorrow's weather, 5% chance of rain means that IF you simulated tomorrow 100 times, in 5 of those times there would be rain. But watch the video if you have time to get a deeper intuitive feel for this.
@@DataDemystified Thanks, you are doing a great job, my friend. Amazing explanation. I know that you have already done a confidence interval video. Are you planning to do more on the same concept with examples?
@@ajaykulkarni576 Thank you! I wasn't immediately planning on it, but what specific topic are you thinking about? Just confidence intervals, or something else?
@@DataDemystified I think topics, where i could see benefiting from your amazing style of teaching, would be 1) interpreting logs (logarithm) very nonintuitive, especially in graphs. 2) An easy layman intuition of confidence interval, how to make sense when someone says 95% confident. 3) Maybe some other statistical terms like R square interpretation, how to interpret a log y within the regression. Your style of teaching is AMAZING .... I have spent so many hours on youtube trying to understand, but your videos have made it so so easy to digest.. you are a great teacher, you may learn so joyful.
@@ajaykulkarni576 Thanks Ajay! I will put those in my list of videos to make for sure!