Type of artificial neural netowrk suitable for learn and then predict forest growth

I'm trying to use an ANN to learn from a large amount of forest measurement data obtained from sampling plots across Ontario, Canada and associated climate data provided by regional climate modelling in this province.

So the following are the inputs to the ANN:

1. Location (GPS coordinates)
2. Measurement year and month
3. Tree species
4. Age
5. Soil type
6. Soil moisture regime
7. Seasonal or monthly average temperature
8. Seasonal or monthly average precipitation
9. More data are available if needed.

And the targets include: (1) Average total tree height and (2) Average tree diameter at breast height

For each sampling plot, the trees have been measured for 1-4 times over 30 years. So my question is what type of ANN can best used to learn from the data and then it can be used for predicting with a set of new input data?

Answer 1

[Vishal Bhutani]

Based on my understanding you want to create an Artificial Neural Network (ANN) for prediction of forest growth. One of the things to start is to use Shallow Neural Network there are two ways to create that:

1.) Use the nprtool GUI

2.) Use a command-line solution

You can find the link for documentation for Classify Patterns with a Shallow Neural Network: https://www.mathworks.com/help/nnet/gs/classify-patterns-with-a-neural-network.html#f9-26645

There are several hyperparameters which you can vary and fine-tune the network like learning rate, number of neurons etc.

Another thing you can explore is Neural Network Toolbox

https://www.mathworks.com/help/nnet/examples/classify-sequence-data-using-lstm-networks.html

[Jiaxin Chen]

Hi Vishal, thanks for providing answers. I've been trying to use shallow neural network to solve the problem, and I think I've been making good progress.

Answer 2

[Greg Heath]

This is aregression prolem, not a classification problem.

 help fitnet

 doc fitnet

Hope this helps

Thank you for formally accepting my answer

Greg

[Jiaxin Chen]

Thank you very much Greg. Yes you're right, this is a regression problem. I've tried different ANN, such as fitnet and cascadeforwardnet. And I've been making good progress.