Naïve Bayes

Setting Up a Naïve Bayes Classifier

Note: Make sure you have your training and test data already vectorized and ready to go before you begin trying to fit the machine learning model to unprepped data.

Load in required libraries

from import NaiveBayes
from import ParamGridBuilder, CrossValidator
from import BinaryClassificationEvaluator
from pyspark.mllib.evaluation import BinaryClassificationMetrics

Initialize Naïve Bayes object

nb = NaiveBayes(labelCol="label", featuresCol="features")

Create a parameter grid for tuning the model

nbparamGrid = (ParamGridBuilder()
               .addGrid(nb.smoothing, [0.0, 0.2, 0.4, 0.6, 0.8, 1.0])

Define how you want the model to be evaluated

nbevaluator = BinaryClassificationEvaluator(rawPredictionCol="rawPrediction")

Define the type of cross-validation you want to perform

# Create 5-fold CrossValidator
nbcv = CrossValidator(estimator = nb,
                      estimatorParamMaps = nbparamGrid,
                      evaluator = nbevaluator,
                      numFolds = 5)

Fit the model to the data

nbcvModel =

Score the testing dataset using your fitted model for evaluation purposes

nbpredictions = nbcvModel.transform(test)

Evaluate the model

print('Accuracy:', lrevaluator.evaluate(lrpredictions))
print('AUC:', BinaryClassificationMetrics(lrpredictions['label','prediction'].rdd).areaUnderROC)
print('PR:', BinaryClassificationMetrics(lrpredictions['label','prediction'].rdd).areaUnderPR)

Note: When you use the CrossValidatorfunction to set up cross-validation of your models, the resulting model object will have all the runs included, but will only use the best model when you interact with the model object using other functions like evaluate or transform.

Last updated