Improving Lithium Battery: Application of Data Mining

In an age controlled by improved technological innovations and ideas there is some increasing gradual depletion in the performance measure of lithium battery which has triggered a shift in its efficiency. This text provides an improvement into lithium battery data life performance prediction using Weka 3.7.1. The classification technique was used to classify the massively extracted dataset from Arbin BT 2000 Battery Testing Repository. However based on the huge size of the dataset 20% which was adequate when dealing with huge data size yielded 10001 instances with 14 attributes. The Multi-Layer Perceptron Sequential Minimal Optimisation and Naïve Bayes were the algorithms used to perform the lithium battery data mining the efficiency. Furthermore the researcher applied k-fold cross-validation with 90% training data and 10% test data which realised Multi-Layer Perceptron 99.6% Sequential Minimal Optimization is 99.7% and Naïve Bayes is 97% with an insignificant error rate.