An empirical comparison between stochastic and deterministic centroid initialisation for K-means variations

Machine Learning - K-Means is one of the most used algorithms for data clustering and the usual clustering method for benchmarking. Despite its wide application it is well-known that it suffers...     

Comparison and analysis of the emissions of a small non-road spark-ignition engine operating under different alcohol–gasoline blended fuels

In this experimental study we focused our interest on comparing the effect of lower and higher molecular mass alcohol–gasoline-blended fuels on the regulated emissions emitted by a small non-road spark-ignition engine. Twenty-one test fuels were used in this experimental study that included gasoline as a reference as well as low and high molecular mass alcohol–gasoline blends containing 5%, 10%, 20% and 40% v/v. In exhaust gases that originated from alcohol gasoline test fuels, low CO/HC and high CO₂/NO_x emissions were observed as the total percentage of alcohol in the blend increased. Methanol–gasoline blends seemed to achieve good combustion efficiency, but the engine will require a catalytic converter against high NO_x emissions. Butanol–gasoline blends in several cases gave lower emissions in comparison with the ethanol and propanol–gasoline blends. Finally, the pentanol–gasoline blends showed exactly the same emission patterns as those of neat gasoline.