CPU and eGPU Support System Based on Naive Bayes Classification

摘要

Central Processing Unit (CPU) and External Graphics Processing Unit (eGPU) known as overclocks which aim to make device exceed the benchmarks set by the device maker. However, until now there has been no determination to rank the two hardware within certain limits, such as the hardware price range. Simple Additive Weighting (SAW) is used to determine CPU rank based on values of Cores, Threads, Base, Clock, and TDP and eGPU based on values of Memory, Bit Rate, GPU Clock, and Memory Clock then on both hardware are grouped based on price. Focus of this research is to test Na?ve Bayes classification algorithm to determine results of the criteria combination between two devices to determine the possible criteria to be "bad" and "good". This classification is used to determine the probability criteria for selecting the combination of CPU and eGPU hardware. Tests carried out on the application of Na?ve Bayes use 80% of training data which has 2776 data and 20% of test data which has 695 data to be tested for accuracy, precision, recall, and F1-score. The results of the tests that have been carried out, the results obtained an accuracy of 0.78, precision 1, Recall 0.764, and F1-Score 0.866.