1. Start the Spark shell:

        $ spark-shell

2. Perform the required imports:

        scala> import org.apache.spark.ml.classification.NaiveBayes
scala> import 
        org.apache.spark.ml.evaluation.MulticlassClassificationEvaluator

3. Load the data into the DataFrame from S3:

        scala> val data = 
        spark.read.format("libsvm").load("s3a://sparkcookbook/patientdata")

4. Split the data into training and test datasets:

        scala> val Array(trainingData, testData) = 
        data.randomSplit(Array(0.7, 0.3))

5. Train the model with the training dataset:

        scala> val model = new NaiveBayes().fit(trainingData)

6. Do the prediction:

        scala> val predictions = model.transform(testData)

7. Evaluate the accuracy:

        scala> val evaluator = new MulticlassClassificationEvaluator()
        .setMetricName("accuracy")
        scala> val accuracy = evaluator.evaluate(predictions)

Here the accuracy is only 55 percent, which shows Naive Bayes is not the best algorithm for this dataset.