Most classification algorithms assume that classes are in a balanced state. However,
datasets with class imbalances are everywhere. The classes of actual medical datasets are …

In recent years, new energy vehicles (NEVs), which are considered to be one of the most
important ways of solving global warming and energy crisis, have seen rapid development …

In ultrasonic metal welding (UMW), tool wear significantly affects the weld quality and tool
maintenance constitutes a substantial part of production cost. Thus, tool condition monitoring …

Microblogs have become a customary news media source in recent times. But as synthetic
text or 'readfakes' scale up the online disinformation operation, unsubstantiated pieces of …

Selecting the correct features when training a classification algorithm, has a significant
impact on the performance of the classifier. More features provide more information, but can …

An innovative low-cost device based on hyperspectral spectroscopy in the near infrared
(NIR) spectral region is proposed for the non-invasive detection of moldy core (MC) in …

Abstract Information overload on Web has been a well-identified challenge which has
amplified with the advent of social web. Good, bad, true, false, useful, useless all kinds of …

Denigration is a specialized form of cyberbullying which describes a recurrent, sustained
and intentional attempt to damage the victim's reputation or ruin the friendships that he or …

This paper proposes a new method that can identify and predict financial fraud among listed
companies based on machine learning. We collected 18,060 transactions and 363 …

Opinion mining plays a major role in text mining applications in consumer attitude detection,
brand and product positioning, customer relationship management, and market research …