This paper proposes an intelligent computational multimodal biometric recognition technique to automatically authenticate face and fingerprint (FP) images using machine learning techniques. The face and FP features are obtained by combining the features from Gabor filter and Convolutional Neural Network (CNN). Then the system performs dimensionality reduction (DR) with the help of Principal Component Analysis (PCA) to avoid the overfitting problem. The outcomes of the proposed system are compared with the existing classifiers say Support Vector Machine (SVM) and Random Forest (RF) concerning recognition rate, precision, recall and F-measure. The result proves that the proposed technique provides the recognition rate of 98.7%, which is higher than the existing state of are classifiers. K-fold cross validation techniques is used in the research work to avoid the overfitting problem.