Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News – A new study on artificial intelligence is now available. According to news reporting outof Dongseo University by News Rx editors, research stated, “Oversampling is often applied as a means towin a better knowledge model.”The news reporters obtained a quote from the research from Dongseo University: “ Several oversamplingmethods based on synthetic instances have been suggested, a nd SMOTE is one of the representativeoversampling methods that can generate syn thetic instances of a minor class. Until now, the oversampleddata has been used conventionally to train machine learning models without statistical analysis, s o itis not certain that the machine learning models will be fine for unseen cas es in the future. However,because such synthetic data is different from the ori ginal data, we may wonder how much it resemblesthe original data so that the ov ersampled data is worth using to train machine learning models. For thispurpose , I conducted this study on a representative dataset called wine data in the UCI machine learningrepository, which is one of the datasets that has been experim ented with by many researchers in researchfor knowledge discovery models. I gen erated synthetic data iteratively using SMOTE, and I comparedthe synthetic data with the original data of wine to see if it was statistically reliable using a box plotand t-test. Moreover, since training a machine learning model by supply ing more high-quality traininginstances increases the probability of obtaining a machine learning model with higher accuracy, it was alsochecked whether a bet ter machine learning model of random forests can be obtained by generating muchmore synthetic data than the original data and using it for training the random forests.”
NETL
NSTL