In this work, we report the detection and discrimination of five commercially available NESTT K-9 explosive compounds, including 2,4,6-trinitrotoluene (TNT), cyclotrimethylenetrinitramine (RDX), pentaerythritol tetranitrate (PETN), potassium nitrate, and potassium chlorate. An array of 48 chemiresistors were fabricated and assembled with four types of gold nanoparticle sensing materials. The functional groups include tetradecylamine (TDA), octadecylamine (ODA), 3-mercaptopropionic acid (MPA), and 4-aminothiophenol (ATP). Machine learning methods were applied to analyze sensor data. The discrimination accuracy of the sensor array was studied over time at various vapor concentrations ( ${p}/{p}_{0}$ ), using several classifiers, including linear discriminant analysis (LDA), support vector machine (SVM), k-nearest neighbors (KNN), and bagged trees (BT), with five-fold cross-validation. More than 90% accuracy was achieved using datasets with more than 2500 sensor measurements. Sensor arrays and classification algorithms exhibit good stability over a range of vapor concentrations. The results demonstrate the utility of machine learning for detection and classification of volatiles, including explosive and related compounds.