How do i use ANN neural network function for predicting for new inputs

I used the NN fitting tool GUI to model a data set of inputs to outputs. How can I use the generated neural network function script to predict output for new set of input data set.

[belkacemi djelloul]

the answer is in this post : https://fr.mathworks.com/matlabcentral/answers/1974584-using-a-trained-ann-model-for-new-set-of-inputs

Good luck

[Emmanuel Emeka]

Thank you so much.

Answer 1

[Shubham]

Hi Emmanuel,

After you've used the Neural Network Fitting Tool (nftool) in MATLAB and generated a neural network function script (let's say the script is named myNeuralNetworkFunction), you can use this function to predict outputs for a new set of input data by following these steps:

1. Load or Prepare Your New Input Data

First, you need to have your new input data ready in a format that the neural network function can process. Typically, your input data should be in a matrix format where each column represents a different input feature, and each row represents a different data point (or vice versa, depending on how you've trained your model).

For example, if you have 3 input features and 10 new data points, your input data matrix (newInputData) should have a size of 10x3.

% Example of preparing new input data
newInputData = [ ... % Your new input data here
                 ];

2. Use the Generated Function to Predict Outputs

Next, you'll use the neural network function script that was generated by the fitting tool. This function takes your input data as an argument and returns the predicted outputs.

% Predict outputs using the generated neural network function
predictedOutputs = myNeuralNetworkFunction(newInputData);

In this step, myNeuralNetworkFunction is the name of the function script generated by the nftool. You should replace myNeuralNetworkFunction with the actual name of your generated function. The newInputData is the matrix of new input data you prepared in the previous step. The function will return predictedOutputs, which contains the neural network's predictions for your new input data.

3. Analyzing the Predicted Outputs

Now that you have the predicted outputs, you can analyze them as needed for your application. This might include comparing the predictions to actual outputs (if you have them), performing further statistical analysis, or using the predictions in a larger application or workflow.

Additional Notes

• Ensure that the new input data is preprocessed in the same way as the training data was. This often includes steps like normalization or scaling based on the parameters used during training.
• If the neural network function script was generated with a separate preprocessing and postprocessing step (sometimes included in the generated script), make sure to apply these steps to your new input data and the predicted outputs accordingly.
• The generated neural network function script should be in your current working directory or in a directory that's on your MATLAB path. If it's not, you'll need to add the directory to your path using the addpath function.

By following these steps, you can use the neural network model you've trained and exported from the Neural Network Fitting Tool to make predictions on new data directly within MATLAB.

[Emmanuel Emeka]

I appreciate your assistance.