My main goal is to use Neural Networks to forecast Sunspot Numbers. Requesting the option of ANN or RNN seems simple enough. However, which is best to learn and utilize for a complete beginner? If there is a GitHub repository for similar Space Science topics based on Neural Networks, please link me to it. I'd be extremely appreciative.
Ashok Silwal Are you performing the right kind of multi-step forecasting? If it is closely related to its neighbours then choose RNN which provides the possibility to model time series dynamic systems and for accuracy you may want to create two different models for each output.
Good luck
can be done very easily.
Split the data 70% or 80% from the beginning and the rest of the data for testing.
Then use
trainFcn = 'trainlm';
hiddenLayerSize = [25];
net = fitnet(hiddenLayerSize,trainFcn);
net.divideParam.trainRatio = 100/100;
net.divideParam.valRatio = 10/100;
net.divideParam.testRatio = 0/100;
[net,tr] = train(net,xtrain,ytrain);
predict_train = net(xtrain)';
predict_test = net(xtest)';
Hi. See the following link:
https://towardsdatascience.com/time-series-forecasting-with-rnns-ff22683bbbb0?gi=f974b2b874e0
Thank you, Pradeep Paraman, for your response. Actually, I have three sets of time-series Earth data and three sets of time-series Solar data. Using Solar data as multiple inputs (say X1, X2 and X3), I need to predict corresponding Earth data (Y1, Y2, Y3). In that case, RNN is best, right?
Abidhan Bardhan, Would you mind recommending a few essential resources for a beginner in this field? It would be quite beneficial. You may have collected the essential resources when you had just started.
- Badges
- Science topic