diff --git a/mllib/src/main/scala/org/apache/spark/ml/ann/Layer.scala b/mllib/src/main/scala/org/apache/spark/ml/ann/Layer.scala
index 45e3da51a6c5d..b5258ff348477 100644
--- a/mllib/src/main/scala/org/apache/spark/ml/ann/Layer.scala
+++ b/mllib/src/main/scala/org/apache/spark/ml/ann/Layer.scala
@@ -776,7 +776,7 @@ private[ml] class FeedForwardTrainer(
 
   // TODO: what if we need to pass random seed?
   private var _weights = topology.getInstance(11L).weights()
-  private var _stackSize = 100
+  private var _stackSize = 128
   private var dataStacker = new DataStacker(_stackSize, inputSize, outputSize)
   private var _gradient: Gradient = new ANNGradient(topology, dataStacker)
   private var _updater: Updater = new ANNUpdater()
diff --git a/mllib/src/main/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifier.scala b/mllib/src/main/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifier.scala
index f92baf41617f0..8cd2103d7d5e6 100644
--- a/mllib/src/main/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifier.scala
+++ b/mllib/src/main/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifier.scala
@@ -88,7 +88,7 @@ private[ml] trait MultilayerPerceptronParams extends PredictorParams
    */
   def setSeed(value: Long): this.type = set(seed, value)
 
-  setDefault(maxIter -> 100, tol -> 1e-4, layers -> Array(1, 1), blockSize -> 100)
+  setDefault(maxIter -> 100, tol -> 1e-4, layers -> Array(1, 1), blockSize -> 128)
 }
 
 /** Label to vector converter. */
@@ -153,8 +153,8 @@ class MultilayerPerceptronClassifier(override val uid: String)
     val data = lpData.map(lp => LabelConverter.encodeLabeledPoint(lp, labels))
     val topology = FeedForwardTopology.multiLayerPerceptron(myLayers, true)
     val FeedForwardTrainer = new FeedForwardTrainer(topology, myLayers(0), myLayers.last)
-    FeedForwardTrainer.LBFGSOptimizer.setConvergenceTol(getTol).setNumIterations(getMaxIter)
-    FeedForwardTrainer.setStackSize(getBlockSize)
+    FeedForwardTrainer.LBFGSOptimizer.setConvergenceTol($(tol)).setNumIterations($(maxIter))
+    FeedForwardTrainer.setStackSize($(blockSize))
     val mlpModel = FeedForwardTrainer.train(data)
     new MultilayerPerceptronClassifierModel(uid, myLayers, mlpModel.weights())
   }
@@ -166,11 +166,11 @@ class MultilayerPerceptronClassifier(override val uid: String)
  * Each layer has sigmoid activation function, output layer has softmax.
  * @param uid uid
  * @param layers array of layer sizes including input and output layers
- * @param weights vector of initial weights for the model
+ * @param weights vector of initial weights for the model that consists of the weights of layers
  * @return prediction model
  */
 @Experimental
-class MultilayerPerceptronClassifierModel private[ml](
+class MultilayerPerceptronClassifierModel private[ml] (
     override val uid: String,
     layers: Array[Int],
     weights: Vector)
diff --git a/mllib/src/main/scala/org/apache/spark/ml/param/params.scala b/mllib/src/main/scala/org/apache/spark/ml/param/params.scala
index 5e1855d6a50b9..cbff804c806fc 100644
--- a/mllib/src/main/scala/org/apache/spark/ml/param/params.scala
+++ b/mllib/src/main/scala/org/apache/spark/ml/param/params.scala
@@ -168,15 +168,8 @@ object ParamValidators {
   }
 
   /** Check that the array length is greater than lowerBound. */
-  def arrayLengthGt[T](lowerBound: Double): T => Boolean = { (value: T) =>
-    val array: Array[_] = value match {
-      case x: Array[_] => x
-      case _ =>
-        // The type should be checked before this is ever called.
-        throw new IllegalArgumentException("Array Param validation failed because" +
-          s" of unexpected input type: ${value.getClass}")
-    }
-    array.length > lowerBound
+  def arrayLengthGt[T](lowerBound: Double): Array[T] => Boolean = { (value: Array[T]) =>
+    value.length > lowerBound
   }
 }
 
diff --git a/mllib/src/test/scala/org/apache/spark/ml/ann/ANNSuite.scala b/mllib/src/test/scala/org/apache/spark/ml/ann/ANNSuite.scala
index 449288a48da65..1292e57d7c01a 100644
--- a/mllib/src/test/scala/org/apache/spark/ml/ann/ANNSuite.scala
+++ b/mllib/src/test/scala/org/apache/spark/ml/ann/ANNSuite.scala
@@ -20,12 +20,14 @@ package org.apache.spark.ml.ann
 import org.apache.spark.SparkFunSuite
 import org.apache.spark.mllib.linalg.Vectors
 import org.apache.spark.mllib.util.MLlibTestSparkContext
+import org.apache.spark.mllib.util.TestingUtils._
+
 
 class ANNSuite extends SparkFunSuite with MLlibTestSparkContext {
 
   // TODO: test for weights comparison with Weka MLP
   test("ANN with Sigmoid learns XOR function with LBFGS optimizer") {
-    val inputs = Array[Array[Double]](
+    val inputs = Array(
       Array(0.0, 0.0),
       Array(0.0, 1.0),
       Array(1.0, 0.0),
@@ -33,7 +35,7 @@ class ANNSuite extends SparkFunSuite with MLlibTestSparkContext {
     )
     val outputs = Array(0.0, 1.0, 1.0, 0.0)
     val data = inputs.zip(outputs).map { case (features, label) =>
-      (Vectors.dense(features), Vectors.dense(Array(label)))
+      (Vectors.dense(features), Vectors.dense(label))
     }
     val rddData = sc.parallelize(data, 1)
     val hiddenLayersTopology = Array(5)
@@ -48,17 +50,19 @@ class ANNSuite extends SparkFunSuite with MLlibTestSparkContext {
     val predictionAndLabels = rddData.map { case (input, label) =>
       (model.predict(input)(0), label(0))
     }.collect()
-    assert(predictionAndLabels.forall { case (p, l) => (math.round(p) - l) == 0})
+    predictionAndLabels.foreach { case (p, l) =>
+      assert(math.round(p) === l)
+    }
   }
 
   test("ANN with SoftMax learns XOR function with 2-bit output and batch GD optimizer") {
-    val inputs = Array[Array[Double]](
+    val inputs = Array(
       Array(0.0, 0.0),
       Array(0.0, 1.0),
       Array(1.0, 0.0),
       Array(1.0, 1.0)
     )
-    val outputs = Array[Array[Double]](
+    val outputs = Array(
       Array(1.0, 0.0),
       Array(0.0, 1.0),
       Array(0.0, 1.0),
@@ -78,8 +82,10 @@ class ANNSuite extends SparkFunSuite with MLlibTestSparkContext {
     trainer.setWeights(initialWeights)
     val model = trainer.train(rddData)
     val predictionAndLabels = rddData.map { case (input, label) =>
-      (model.predict(input).toArray.map(math.round(_)), label.toArray)
+      (model.predict(input), label)
     }.collect()
-    assert(predictionAndLabels.forall { case (p, l) => p.deep == l.deep})
+    predictionAndLabels.foreach { case (p, l) =>
+      assert(p ~== l absTol 0.5)
+    }
   }
 }
diff --git a/mllib/src/test/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifierSuite.scala b/mllib/src/test/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifierSuite.scala
index a42b0b362345f..ddc948f65df45 100644
--- a/mllib/src/test/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifierSuite.scala
+++ b/mllib/src/test/scala/org/apache/spark/ml/classification/MultilayerPerceptronClassifierSuite.scala
@@ -45,7 +45,8 @@ class MultilayerPerceptronClassifierSuite extends SparkFunSuite with MLlibTestSp
     val result = model.transform(dataFrame)
     val predictionAndLabels = result.select("prediction", "label").collect()
     predictionAndLabels.foreach { case Row(p: Double, l: Double) =>
-      assert(p == l) }
+      assert(p == l)
+    }
   }
 
   // TODO: implement a more rigorous test