Hi, thank you very much for sharing this video :) the only one tutorial that I was able to follow. One question, my predicted variable is not categorial, but it´s an area of deforestation. So, Can I use the code you shared in this video?
Hi, thank you for share this great tutorial, just one question, this code that herein you provide for us ¿is sufficient to work with imbalanced data sets? Or contrary, ¿do you consider that I need to apply some changes? If your answer is yes for the second question, ¿what would these be? Thanks in advance for your reply
i got an error message when creating the random forest model. it says "Error in randomForest(Inactive_flag ~ ., data = training) : could not find function "randomForest". i used the inactive flag in my dataset, but i want to know i could not find the function randomForest. i already installed the randomForest package and loaded the library and it works fine
Thanks for this. One question: why do you sample with replacement for your training/test set? That means you will pick some of the values more than once, right?
Hi thank you so much for your comment. Whilst sample replacement means picking values more than once in this case as we usually use the set.seed command where one particular observation will be only assigned to EITHER TRAINING OR TESTING. If you print the index vector you will find that. I have explained about set seed function in another video having said that I have forgotten to included it in this video. But yes I guess whenever I have used the sample function for statistical purposes I have always kept sample replacement bro be true (possibly due to force of habit). Rest assured in this case one observation will either be assigned to training or testing because of two fail safe protocols 1. the number of rows in the sample function is equal to the number of rows in my data 2. The aforementioned set seed function will ensure no matter how many times we run the code once an observation is assigned to either training or testing it remains so. Thank you once again I will add set.seed in my upcoming ML videos. Please take a look at my Regression Model in R video where I have printed the index function
Very well explained! simple and to the points. Thank you very much. Just one question, if you have more than 53 factor variable how do you perform RF? TIA
Firstly Thank you so much for your kind words :) and to answer your query - You can but it would be better to use PCA or some other dimension reduction technique before :).
EXCELLENT tutorial I must say.... you are born with extraordinary God-gifted abilities my dear.
Man, you must come back to channel and continue to teach us. YOU MUST.
Quickest and best random forest video out there thanks for this!!
Made it so simple and illustrative... thanks a lot
Wow, you made it look so easy. Subscribed!
A super helpful video for beginner like me. plz, keep going on. most underrated channel in ML.
what an awesome video. thanks very much
how can incorporate upsamling or downsampling before running the random forest model? neeeeed help pls
You are the best!!!
Hi, what is this "predict(...)" function? Is it from 'randomForest" or it's R built-in function? Thanks
Hi, thank you very much for sharing this video :) the only one tutorial that I was able to follow. One question, my predicted variable is not categorial, but it´s an area of deforestation. So, Can I use the code you shared in this video?
Very helpful thank you!
your videos are super helpful! Keep it up!
Thanks :)
Hi, thank you for share this great tutorial, just one question, this code that herein you provide for us ¿is sufficient to work with imbalanced data sets? Or contrary, ¿do you consider that I need to apply some changes? If your answer is yes for the second question, ¿what would these be? Thanks in advance for your reply
i got an error message when creating the random forest model. it says "Error in randomForest(Inactive_flag ~ ., data = training) :
could not find function "randomForest". i used the inactive flag in my dataset, but i want to know i could not find the function randomForest. i already installed the randomForest package and loaded the library and it works fine
Does your `training` object contain the `Inactive_flag` column?
Thanks for this. One question: why do you sample with replacement for your training/test set? That means you will pick some of the values more than once, right?
Hi thank you so much for your comment. Whilst sample replacement means picking values more than once in this case as we usually use the set.seed command where one particular observation will be only assigned to EITHER TRAINING OR TESTING. If you print the index vector you will find that. I have explained about set seed function in another video having said that I have forgotten to included it in this video. But yes I guess whenever I have used the sample function for statistical purposes I have always kept sample replacement bro be true (possibly due to force of habit). Rest assured in this case one observation will either be assigned to training or testing because of two fail safe protocols 1. the number of rows in the sample function is equal to the number of rows in my data 2. The aforementioned set seed function will ensure no matter how many times we run the code once an observation is assigned to either training or testing it remains so. Thank you once again I will add set.seed in my upcoming ML videos. Please take a look at my Regression Model in R video where I have printed the index function
浓郁的咖喱味道
Very well explained! simple and to the points. Thank you very much. Just one question, if you have more than 53 factor variable how do you perform RF? TIA
Firstly Thank you so much for your kind words :)
and to answer your query - You can but it would be better to use PCA or some other dimension reduction technique before :).
I will pay someone to find me an R tutorial that focuses on continuous variables instead of factors/classifications.
Thank you, this was very helpful!
Thanks :)
What is the version of your R studio?
why are you angry 🤣
im getting 90% accurarcy
Good video, you will have to my subscription.
Thanks!!!