Trained svm model can't make predictions with new data

load('Model.mat');
new_data = [5.1, 3.5, 1.4, 0.2]; 
prediction = predict(Model, new_data);

matlab2017b，What can I do to make it right?thanks

Answer 1

[Sulaymon Eshkabilov]

Presuming that you are working with fitrsvm() for regression model development, e.g.,

% E.g. Data for model training:
x_train = reshape(1:20, [10,2]);
y_train = (1:10)';
% Train regression SVM model:
SVM_model = fitrsvm(x_train, y_train);
% Predictor test data:
x_test = [20, 21; 24, 25];
% Make predictions using the trained model
y_pred = predict(SVM_model, x_test)

y_pred = 2×1

14.6763 18.3469

%% OR load the existing model (MODEL_SVM.mat contains the model called Model_SVM): 
load('MODEL_SVM.mat')
y_pred2 = predict(Model_SVM, x_test)

y_pred2 = 2×1

14.6763 18.3469