Aim: Prediction of Employee Attrition Using Novel K Neighbors Classifier Algorithm Compared with Logistic Regression Algorithm to Increase Accuracy. This information may be used to predict employee turnover in organizations and its contributing variables. In this work the two algorithms are logistic regression algorithm and the Novel K Neighbors Classifier. With the Clincalc.com website, the sample size is determined while taking the 0.05 threshold, 95% CI, and 80% g power into consideration. The mean difference and standard error difference from the accuracy is 12.450 and 0.76149 with a significance level of 0.13. Hence the significant value from equal variance assumed is p = 0.13 which is less than Novel K Neighbors Classifier Algorithm and Logistic Regression. Here the 95% confidence interval has been divided into two types: the lower is 10.9084 and the upper is 13.99156. The sample size per group is 1444 for both the current method and the recommended methodology. Before being examined and tested in the Google Colab programming environment, the dataset is scaled down via slicing. The testing dataset 30% of actual dataset is dataset which is 441 and training dataset is 70% of actual dataset is 1029. Results: It has been analyzed that the proposed system Novel K Neighbors Classifier got the accuracy of 89.4 when compared with the existing system Logistic Regression Algorithm which got the accuracy of 75.8% of sample size of 1444. Conclusion: After implementing the Novel K Neighbors Classifier it has more accuracy than the existing system Logistic Regression.