Applying Machine Learning in the Prediction and Identification of Children Affected by Prenatal Alcohol Exposure