Accurately quantifying the goodness of music based on the seemingly subjective taste of the public is a multi-million industry. Recording companies can make sound decisions on which songs or artists to prioritize if accurate forecasting is achieved. We extract 56 single-valued musical features (e.g. pitch and tempo) from 380 Original Pilipino Music (OPM) songs (190 are hit songs) released from 2004 to 2006. Based on an effect size criterion which measures a variable's discriminating power, the 20 highest ranked features are fed to a classifier tasked to predict hit songs. We show that regardless of musical genre, a trained feed-forward neural network (NN) can predict potential hit songs with an average accuracy of Φ NN = 81%. The accuracy is about +20% higher than those of standard classifiers such as linear discriminant analysis (LDA, Φ LDA = 61%) and classification and regression trees (CART, Φ CART = 57%). Both LDA and CART are above the proportional chance criterion (PCC, Φ PCC = 50%) but are slightly below the suggested acceptable classifier requirement of 1.25*Φ PCC = 63%. Utilizing a similar procedure, we demonstrate that different genres (ballad, alternative rock or rock) of OPM songs can be automatically classified with near perfect accuracy using LDA or NN but only around 77% using CART.

Penmanship has a high degree of uniqueness as exemplified by the standard use of hand signature as identifier in contract validations and property ownerships. In this work, we demonstrate that the distinctiveness of one's writing patterns is possibly embedded in the molding of chalk tips. Using conventional photometric stereo method, the three-dimensional surface features of blackboard chalk tips used in Math and Physics lectures are microscopically resolved. Principal component analysis (PCA) and neural networks (NN) are then combined in identifying the chalk user based on the extracted topography. We show that NN approach applied to eight lecturers allow average classification accuracy (Φ NN ) equal to 100% and 71.5 ± 2.7% for the training and test sets, respectively. Test sets are chalks not seen previously by the trained NN and represent 25% or 93 of the 368 chalk samples used. We note that the NN test set prediction is more than five-fold higher than the proportional chance criterion (PCC, Φ PCC = 12.9%), strongly hinting to a high degree of unique correlation between the user's hand strokes and the chalk tip features. The result of NN is also about three-fold better than the standard methods of linear discriminant analysis (LDA, Φ LDA = 27.0 ± 4.2%) or classification and regression trees (CART, Φ CART = 17.3 ± 3.7%). While the procedure discussed is far from becoming a practical biometric tool, our work offers a fundamental perspective to the extent on which the uniqueness of hand strokes of humans can be exhibited.