# How to use random forest method

55 views (last 30 days)
Kanakaiah Jakkula on 2 Mar 2018
Commented: Jan Startek on 17 Mar 2021
Hi, Below is my training data (v1,v2,v3 are process variables, and Y is the response variable,
Based on training data, given set of new v1,v2,v3, and predict Y. I want to make prediction using "Random forest tree bag" (decisiotn tree regression) method.
Training data:
v1 v2 v3 Y
16.00 21.05 25.01 8
14.44 22.79 27.02 1
14.69 21.83 25.10 1
15.32 21.09 28.35 4
15.22 23.38 28.11 2
14.6 22.77 27.8 3
15.52 22.33 27.78 0
15.30 22.28 28.18 0
15.47 20.94 28.14 5
14.04 23.41 25.22 9
15.67 20.89 28.97 8
14.14 20.73 26.22 6
14.80 20.87 25.82 2
14.63 22.33 27.03 0
15.66 22.22 28.22 1
14.40 20.73 28.13 7
14.09 21.04 26.25 5
14.61 23.48 26.24 4
14.63 22.33 27.03 0
15.53 22.79 28.37 6
14.51 22.78 28.8 5
15.76 23.27 26.62 3
15.42 21.68 27.59 0
Make prediction for below two set of test data, what Y
test data:
v1 v2 v3
15.29 21.39 27.07
14.53 23.18 26.31

Bernhard Suhm on 2 Apr 2018
You could read your data into the Classification Learner app (New Session - from File), and then train a "Bagged Tree" on it (that's how we refer to random forests). However, given how small this data set is, the performance will be terrible.
##### 3 CommentsShow 1 older commentHide 1 older comment
Javaid Iqbal on 4 Jul 2018
Sir may you please help me regarding SVM classification. When I use predict test data I'm unable predict thr right solutions.
yfit=trainedModel.predictFcn(T);
I got result yfit=0 as predicted but when it should give yfit=1 result for tru value but this showing same yfit=0
Jan Startek on 17 Mar 2021
Does Bagged Tree really use a ranfom forest algorithm ( Breiman random forest? ). My point is I don't see where in the code a random variable selection for each node in grown trees is perforemd. TreeBagger (https://se.mathworks.com/help/stats/treebagger.html#bvf7_tc-1) seems to have such an option with the use of "'NumPredictorsToSample'" however I can't find an analogus option in TreeBagger.
My question boils down to, is there a random variable selection performed at each split in the grown trees in the Bagged Trees algorithm in clasifiactionLearner?

### Categories

Find more on Classification Ensembles in Help Center and File Exchange

### Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!