Statistics 101: Multiple Linear Regression, Evaluating Basic Models Continued
HTML-код
- Опубликовано: 21 авг 2024
- In Part B of this video, we learn about how to evaluate basic multiple regression models including variable selection, and how to assess the impact of problem variables on the overall model. We will examine two-variable and three-variable models and then put those together with the models from Part A to choose the best one for making predictions.
My playlist table of contents, Video Companion Guide PDF documents, and file downloads can be found on my website: www.bcfoltz.com
#statistics #regression #machinelearning
I honestly don't believe there is any other better Statistics teacher in the world than you. One can just also argue that with the fact that, from the video, it's completely apparent that you make these videos out of a sheer passion for statistics. Hats off man. God bless you.
Thank you for these videos man. Thank you for deciding to make them, you've made such a difference for me. I honestly understand multiple regression so much clearer now. Please keep up the good work, RUclips needs you, as do many many students around the world!! Love from Ormskirk, England
Masterpiece. Veritable and digestible 101 presentation
AGREE!!
Incredible, incredible videos. You made multiple regression look like ABCD. Outstanding
Thank you so much for these videos Brandon, I don't post on youtube ever but just had to say thank you for making this very confusing area of maths much more intuitive and fun to learn! I have been slowly going through all the Statistics 101 playlists in order and I finally feel like I am understanding these concepts individually and how they relate to each other.
Brandon, like many others who have commented previously I can only reiterate their praise for your brilliant teaching methods. You are, no exaggeration, a hero for providing these tutorials for free without which I think my brain would have definitely melted :) Thank you
You made me excited for statistics. I didn't even think that was possible. Thank you so much for your videos! I have been struggling all semester to understand these concepts, and you make it seem so easy and understandable. Kudos and please continue what you are doing! You are helping so much people. :)
I am listening to you instead of the professor I am taking this class. Since I don't know the basics, your channel is a lifesaver to me.
Brandon - I have found all your videos extremely helpful both from an academic as well as practical application perspective. You make these "could be extremely confusing concepts" very easy to understand - so thank you. I have subscribed and donated on gofundme. Keep up the good work - looking forward to more from you.
Seriously the best videos on stats on RUclips. Brandon Foltz is a world class educator
Thank you so very much. I had no idea of stats before this and was trying my absolute best to do my thesis by reading and watching spss tutorials, but finally your videos made me feel so much more clear in my head. You've made a huge difference. Thank you.
my God..... thanks very much....you just made me to work in the field of research with full confidence.... great work.... please keep doing such incredible work
I'm getting an A in my biostatistics class because of you for my graduate program. Thank you so much for posting these! Not all heroes wear capes
Thank you, Brandon, I am a master's student in the Netherlands struggling with research-intensive courses. I have failed one of the research courses and will have a second attempt in July 2020, I am hoping that after watching your videos, I will have a better grasp of how to interpret research results and how to write a better master's thesis
You did a great job of choosing an example that showcases that the simplest model can often be the most accurate/representative, while displaying notions of problems associated to multicollinearity and overfitting!
Haven't really seen a better professor than you Brandon.. Hats off for the way you teach ,the way you talk and the way you enlighten me with the knowledge
Mr.Brandon....one word enough to u....YOU ARE "A W E S O M E" ....and u r god to stat students.
just awesome explanation, make the complex subject simple! that's the best teaching!!!
So glad I found your videos. They are extremely helpful! You explain things so clearly and the examples you use are relatable. Thank you so much!
I genuinely appreciate you Mr.Brandon. You are really helping me to finish my thesis. Thanks
"VIF above 10 means terminally death problems", never had such a clear explanation about VIF, thanks for your hard work!
really AMAZING video! Even a high school student can understand comprehensively.
Thanks you Brandon for taking out your time and providing such beautiful lectures. U are awesome.
no one else can explain this better than you. Thanks much, what you are doing is such a great thing !!
thanks a lot, good at explaining multiple times so I don't forget this in like 10 min, your a good teacher!
Multiple regression never seemed so simple. Thanks a lot for such a detailed explanation Brandon
Thanks for all these brilliant videos Brandon! I've majored in math&stat and now I've been worked for many years. Reviewing the stat knowledge from your videos makes me better understand some concepts than at university then. I've never had such clear clue about how to review a model like now. Thanks Again!
Really it is more than very clearly explained . Many thanks prof. Dr Issam m.a.shakir - Iraq-baghdad
I know you know, but these videos are just so good!!!! You are very knowledgeable, thank you for sharing it with us!
Safe to say I think your videos have just saved one of my grades towards my degree, thank you so much Brandon!!! :)
respect to Brandon for making multiple regression seems so easy and so much more enjoyable that the class my parents paid for.
I never understand my lecturer did same thing but u explain really well now i understand well. u clear each and every factors and significance of that variable first time i have seen educational video 1 and half hour continiously.
Thanks a lot Brandon for your way of understanding...
These videos are just really well done - thanks for all the effort you've put in Brandon!
Thanks so much for these videos. I say a prayer for you every time your videos make my eyes widen in comprehension. This is such a lifesaver!
Hi Brandon, Your presentation is awesome! You've helped solve the problem that I'm having a headache with. Your videos deserve billions of thumb-ups. Thanks!
These videos have become my standard reference resource for statistics
Thanks too much Mr. Brandon, your videos have made a significant improvement for me in understanding econometric, I highly appreciate your efforts, please keep on helping us to empower ourselves, we really need you. God bless you.
I really appreciate your video. I'm studying stats by myself with this series! I totally know it's off the track but want to comment one thing. I think it DOES make sense that if milesTraveled is held constant, then travelTime is expected to decrease when gasPrice go up. Because if you accelerate more, u need more gas then you accelerate less.(assuming you are traveling the same distance) Obviously, more acceleration reduces travelTime. I understood y x1, x3 regression is pointless. Just wanted to say the minus coefficient might make sense in a real world.
Sorry for my coarse English. I'm not native so I wonder if I used right words for the points I want to make.
I thought the same thing.
i very much agree with you.
man it's so helpful, make me love the multiple regression deeply!!! Appreciate that!!
You're amazing! The video just flew by and I felt like I understood literally everything due to your way of instructing! You have a gift, sir!
huge thanks to you for making the video... i cant say how much i am thankful to you...
Love from Nepal
Thank you so much Brandon for your videos. You explained every single point and make the stats be much clearer for me. They are very helpful for my assignment. Such exceptional lectures!
Just wonderful! I followed the lecture using R and replicating the results, THANK YOU SO MUCH!!!!
Wow that was an awesome lecture. Now I understand model selection!!
Thank you so much!! I was struggling but your videos have really helped. You are the real MVP Brandon.
So much appreciation for the time and patience you've put into these videos.
Thank you. Your videos are very clear to understand. :)
One hypothesis on why travel time might go down due to increase in gas prices is that there might be fewer cars on the road.
A wonderful lesson! I learnt everything in the matter of minutes which otherwise would take me months
Wow, your video is wonderful. I feel frustrated in other class about linear regression, so I turn here for some help. You really help me! All the concepts are very clear. I have basic important ideas about multiple regression now.
You're an absolute legend Brandon, thanks for your fantastic teaching!!!
Thank you very mucho for these awesome videos Brandon. You're the best!
I am soooo happy to find you :D literraly way better than my statistic professor
These videos are making my life soooooo much easier !!!!! Thank you very much, Bradon !!!
Brandon, YOU ARE THE BEST!
Brandon Foltz, your presentations are awesome and very well explained in simplest way.
LOVE your videos. They have really helped me a lot. Very detailed and straightforward! !
I watched the series starting from video 1. Very helpful.
You are an excellent teacher. I like your videos very much. Very helpful video to select the best regression model. Thank you very much
This helps my thesis! thanks so much
I’ve just been schooled! Thank you. Will be bookmarking this.....
Thank you for this video. you are a great teacher.
I love you!!! There are no words to describe how good you are at teaching!
Thank you so much Brandon, you made it possible for me to understand the regression. God bless you brother
You made it easy peasy. Thank you so much!
Hi Brandon, great tutorial video here. A million thanks for the upload. I just have a question on your interpretation regarding relevance of gas price to time travelled. I think in some context this is meaningful. For example, I could imagine that if gas price is high, drivers would maintain high speed to save gas or choose expressways or tollways to avoid traffic. The logic for avoiding traffic is that it increases gas consumption. Hence, the result of these strategies would be lesser time of travel.
Would be very glad to hear your idea on this.
Dexter Pante Hi Dexter! That relationship is definitely something that could be explored further if gasPrice had a linear relationship with travelTime to begin with, but in this case it does not. There is no pattern. But you bring up a great point in that when interpreting regression coefficients there is an element of judgement to it. In this example we are using Google to determine the best route based on distance/traffic/location so the route is somewhat predetermined. It would depend on how often the vehicle is being operated near its most fuel efficient speed as well. Going faster to save time usually uses more fuel than the same amount of fuel over the time span that was saved. Cool stuff! :)
+Brandon Foltz You're videos are great, so first, thank you. Just one slight suggestion that may have helped in this scenario. It also took me a few minutes to understand why you excluded the model with gasPrice. In real life, we may get something unexpected that produces an "odd" coefficient, so I was unclear why this would automatically exclude the model. I kept thinking, "although it isn't what we expect, it could be a legitimate finding". However, I realized after I went back to my notes on your previous videos that you covered in your second video that gasPrice was not correlated with travelTime and should have never been included in the model in the first place; you only included it for learning purposes. I think many of us just needed a reminder of that fact in this video (just brief showing of the correlation slide from video 2 that showed no correlation on the scatterplot and/or the Pearson correlation p value).
These are amazing...thank you thank you thank you! I have learned statistics so long ago and needed a refresher course...was struggling with the textbooks...your real life examples make everything so simple...thank you again...
Another B.F. Masterpiece!
Brandon, you saved my thesis. Thank you.
Brandon. Thank you for these videos. You are a blessing my friend!!
Thanks for this series! You make such an amazing teacher.
Not all heroes wear capes. Great work!
This is the best series of videos; Thanks a lot!!!
Your videos helped me a great deal! Thank you!
I appreciate you. These videos are really very helpful. Thanks alot.
thank you so much Brandon, this is really good work and it is easy to understand. I appreciate that
Excellent videos. Love them. Thank you so much
Great videos! My hypothesis for the relationship between gas price and travel time would've been that lower gas prices --> more travel time (since I'd be more likely to stop for gas if the price is really low), but I don't think I'd expect that relationship to be linear since only very low prices might sway me. Maybe that's just how I think though haha
I was so lost in class.
Your video is soooooooooooo helpful
Thank you so much
long, but definitely worth the time! awesome again! thank you for doing this!
Incredible thank you Brandon! You made my life easier.
Thank you for saving my future
YOU ARE A LIFE SAVER
Hello Brandon,
Exceptional lectures! Thank you very much for making it available to all of us. Finally, statistics make sense to me. Would you please make these lectures available in a book format? Majority of statistic books are hard to understand or use. Your instructions are extremely helpful and practical. Even if you make it available in a pdf format, I would love to purchase it. If you don't plan on writing a book, could you please give us some, book, references?
Once again, I appreciate your hard work and making it possible to learn something very useful.
Brandon thanks so much for making this sooooo clear!!! I wish you were my professor !!!!
Brandon you are amazing! Thank you for your work! I enjoyed it on a Sunday evening..
+QueenDolcinea Oh thank you so much! I have been blessed with wonderful teachers my entire life.
Your videos are always so thorough and helpful. It's perfect how you take the time to explain it fully! And also take the time to go over obscure but very important side bars!!! Just one comment- the negative coefficient for travel time and gas prices makes ense to me- if gas prices go up drivers are less apt to joy ride or take the long road home ;D
Excellent one as always
Excellent series in deed!
Thank you Sir so much.
That was not long, that was amazing
Keep up the good work, your videos are the best... Cheers from Bosnia :)
Very thorough and a great help!! Thanks.
Great Going!!! Priceless contribution.
Great tutorial. Thank you very much
This help SO much - thanks a 1000 !!!
Hello Brandon. An excellent video tutorial. Really helped alot. Thank you
Really and extremely helpful Brandon. Thank you very much indeed.
Hi Brandon,
Thank You very much for sharing your knowledge by making these painstaking videos.
I saw your videos about multiple regression. Here the two effective indipendent variables are correlated. And so not advisable to use. But what when the two effective variables are not correlated. For example,
Let's check start time of the trip for each trip. It is not correlated to number of delivery or miles but the sooner one start, the less traffic they face and thus delivery time reduce if start time is early.
In such a case I would like to know if we should use two indipendent variable for regression or one?
2.
When there are more delivery in a long trip they are correlated but at times there are more delivery even in short distance in such cases those indipendent variables add value to the model. I mean the independent variables are not constantly correlated and that matters. How to make use of such knowledge in modeling?
Will appreciate if you share calculation formula of p value, R, R2, S etc when you run the regression. Having formula in front of eyes reminds definitions easily while evaluating the value of the calculated matric.
Thanks.
But
Thank you very much for making these videos on stats. Clear,Concise and Simple. I was also looking for advanced topics where we check AIC and BIC. Also, how do we fix the variables when the assumptions are violated? Is there already a video out there? Appreciate your help. Thanks once again.
Thank you so much for sharing the video !
So interesting 1 Thank you brother.