R Studio doubt : I am building a predictive model with 1 million observations and having 15 variables .i am getting error like -" Can not allocate the vector of 432GB " or " Can not allocate the vector of 3.8 GB " I am using 16GB RAM .my file size is just 140MB . and i closed all the applications in my system .still error remains same . Any suggestions much appreciated..
Is there any way that the classification tree can provide information on which variables will provide more accurate classification among given variables ?
Good tutorial. About just a phrase at the end, you should choose the model with less MSE on the test set (just to make it clear), because if not, you could be overfitting your data and your prediction power will be worthless. It could be also interesting analyse the cost of complexity pruning in this example, maybe with more observations... Thanks!
Thank you Juan, I specifically went to the comments to check if this correction had been made! Especially since order became the criteria for my first split.
jalayer hi. first i would like to thank you for sharing your knowledge with us. please can you explain on what logic algorithm does the decision trees base on ? and second - can you make for us video about random forest algorithm as well ? (in R thank you so much
I have a question: Where are nodes 4,5,8,9, etc.? In the output we can only see nodes 1,2,3,6,12,13,26,27,7,14,15. Can anyone help me with this question?
Nodes 4,5,8,9 are dropped from the model. The output is an optimal subtree , found by adding an additional parameter tothe objective function which weighs the cost of an additional tree against its benefit in reduction of your SSE/MAPE/MAE
In case anyone else has this question going forward, the type is a just the style and a personal preference (or predicted audience preference). I personally prefer type = 2 for this chart. It ranges from 0 to 4.
Thank you so much for such a comprehensive tutorial! You've helped me finish my graduation paper!:)
Thank you Jalayer for creating this Video.
Great video!
Nice choice of data, I need sleep
Thank you! This video helped me in writing a term paper on data analysis.
Thank you for your very clear example. Regards,
I’m confused. Why is the method a nova if it’s a CART. Doesn’t CART use gini?
Very nice video.. Thanks for the explanation! :)
Based on the background noise, I can be hundred percent certain that you are from NY.
hi
can you please tell me how to increase the accuracy of decision tree?
thanks in advance..
Wow you made this look so easy in R. I need to figure it out in Matlab and have my own code written for a class.
R Studio doubt :
I am building a predictive model with 1 million observations and having 15 variables .i am getting error like -" Can not allocate the vector of 432GB "
or " Can not allocate the vector of 3.8 GB "
I am using 16GB RAM .my file size is just 140MB . and i closed all the applications in my system .still error remains same .
Any suggestions much appreciated..
Awesome!!!!
Amazing Video & Thank You- I’ve spent hours on a simple regression tree- but thanks to you I’m free!!!
Crystal clear explanations. Excellent video.
Is there any way that the classification tree can provide information on which variables will provide more accurate classification among given variables ?
Yeah! Maybe use summary(rpart model)..summary(m1) in case of this tutorial and then scroll down to "variable importance". Hope this helps.
Good tutorial. About just a phrase at the end, you should choose the model with less MSE on the test set (just to make it clear), because if not, you could be overfitting your data and your prediction power will be worthless.
It could be also interesting analyse the cost of complexity pruning in this example, maybe with more observations...
Thanks!
database is df
Thank you Juan, I specifically went to the comments to check if this correction had been made! Especially since order became the criteria for my first split.
Thank you very much!
jalayer hi. first i would like to thank you for sharing your knowledge with us. please can you explain on what logic algorithm does the decision trees base on ?
and second - can you make for us video about random forest algorithm as well ? (in R
thank you so much
Incredible explanation, thanks
Thanks. Very clear explanation
thank you a lot
super and great
I do not have enough words to thank you !
I have a question: Where are nodes 4,5,8,9, etc.?
In the output we can only see nodes 1,2,3,6,12,13,26,27,7,14,15.
Can anyone help me with this question?
Nodes 4,5,8,9 are dropped from the model. The output is an optimal subtree , found by adding an additional parameter tothe objective function which weighs the cost of an additional tree against its benefit in reduction of your SSE/MAPE/MAE
can we get the dataset that you are using
thanks so much! very helpful.
How to implement regression tree (CART) in python ? Please help me, thank you
fire ass video
How did you chose type=3?
In case anyone else has this question going forward, the type is a just the style and a personal preference (or predicted audience preference). I personally prefer type = 2 for this chart. It ranges from 0 to 4.
¿where did P1 come from?.
I didn't find it calculated in this video