Dear Avinash,

I have extensively used decision tree method during my doctoral research. Check these two papers in my profile:

1) Landscape control of nitrous oxide emissions during the transition from conservation reserve program to perennial grasses for bioenergy

2) Designing efficient nitrous oxide sampling strategies in agroecosystems using simulation models

I believe you have NDVI as a response variable and you want to predict that with lat-long and possibly other variables. I have used "rpart" function in R to construct the tree. I would be happy to help you with that. Before that, I need to know more about what exactly you want to do and the data structure.

I hope it was helpful.

Best,

Debasish

Dear Debasish

Thanks for your warm response. Actually, i have some NDVI value of particular crops (i.e wheat, gram , mustard) extracted with the help of GPS point (GPS point was taken during field visit). Now i want to classify various crops with the help of these NDVI value of specific crop.

Regards with respect

Avinash

I want to use these value in decision tree method to classify various crop. 

Hi Avinash,

If you are familiar with R statistical software, you can try something like this:

#install the required packages

library(rpart)

set.seed(500) # Use the same seed to get the same forest output

# Make the tree

tree=rpart(NDVI~Crop, data=data, method="anova", control=rpart.control(minsplit=10))

#plot the tree

plot(tree, margin=0.1, uniform=T, main="")

text(tree, use.n=T, cex=0.9)

summary(tree)

This should make a tree classifying the crops based on their respective NDVI values. You can also include other independent variables if you have measurements available.

Let me know if this works.

Good luck!

Thanks debasish sir

you can read ecognition user manual to do what you want.