diff --git a/rumale-neural_network/lib/rumale/neural_network/base_rbf.rb b/rumale-neural_network/lib/rumale/neural_network/base_rbf.rb
index ba6af5e8..633689c4 100644
--- a/rumale-neural_network/lib/rumale/neural_network/base_rbf.rb
+++ b/rumale-neural_network/lib/rumale/neural_network/base_rbf.rb
@@ -64,7 +64,7 @@ def find_centers(x)
         # initialize centers randomly
         n_samples = x.shape[0]
         sub_rng = @rng.dup
-        rand_id = Array(0...n_samples).sample(n_centers, random: sub_rng)
+        rand_id = Array.new(n_centers) { |_v| sub_rng.rand(0...n_samples) }
         @centers = x[rand_id, true].dup
 
         # find centers
diff --git a/rumale-neural_network/spec/rumale/neural_network/rbf_regressor_spec.rb b/rumale-neural_network/spec/rumale/neural_network/rbf_regressor_spec.rb
index 1dd9eb55..7eda5f31 100644
--- a/rumale-neural_network/spec/rumale/neural_network/rbf_regressor_spec.rb
+++ b/rumale-neural_network/spec/rumale/neural_network/rbf_regressor_spec.rb
@@ -12,29 +12,62 @@
   let(:n_samples) { x.shape[0] }
   let(:n_features) { x.shape[1] }
   let(:n_outputs) { y.shape[1] }
-  let(:estimator) { described_class.new(hidden_units: 64, reg_param: 1e4, random_seed: 1).fit(x, y) }
+  let(:hidden_units) { 64 }
+  let(:estimator) { described_class.new(hidden_units: hidden_units, reg_param: 1e4, random_seed: 1) }
   let(:predicted) { estimator.predict(x) }
   let(:score) { estimator.score(x, y) }
 
   shared_examples 'regression' do
+    before { estimator.fit(x, y) }
+
     it 'fits model for given dataset.', :aggregate_failures do
       expect(predicted).to be_a(Numo::DFloat)
       expect(predicted).to be_contiguous
       expect(predicted.ndim).to eq(y.ndim)
       expect(predicted.shape[0]).to eq(n_samples)
+      expect(estimator.centers).to be_a(Numo::DFloat)
+      expect(estimator.centers).to be_contiguous
+      expect(estimator.centers.ndim).to eq(2)
+      expect(estimator.centers.shape[0]).to eq(hidden_units)
+      expect(estimator.centers.shape[1]).to eq(n_features)
+      expect(estimator.weight_vec).to be_a(Numo::DFloat)
+      expect(estimator.weight_vec).to be_contiguous
+      expect(estimator.weight_vec.ndim).to eq(2)
+      expect(estimator.weight_vec.shape[0]).to eq(hidden_units)
+      expect(estimator.weight_vec.shape[1]).to eq(n_outputs)
       expect(score).to be > 0.98
     end
   end
 
-  context 'when single regression problem' do
-    let(:y) { single_target }
+  context 'when the number of hidden units is less than the number of samples' do
+    context 'when single regression problem' do
+      let(:y) { single_target }
+      let(:n_outputs) { 1 }
+
+      it_behaves_like 'regression'
+    end
 
-    it_behaves_like 'regression'
+    context 'when multiple regression problem' do
+      let(:y) { multi_target }
+
+      it_behaves_like 'regression'
+    end
   end
 
-  context 'when multiple regression problem' do
-    let(:y) { multi_target }
+  context 'when the number of hidden units is greater than the number of samples' do
+    let(:hidden_units) { 400 }
 
-    it_behaves_like 'regression'
+    context 'when single regression problem' do
+      let(:y) { single_target }
+      let(:n_outputs) { 1 }
+
+      it_behaves_like 'regression'
+    end
+
+    context 'when multiple regression problem' do
+      let(:y) { multi_target }
+
+      it_behaves_like 'regression'
+    end
   end
 end