Analysis of Stock Market Values Using an Optimal Probability with KNN

The current trend for saving an individual's economic trouble amid a pandemic is stock prediction. Most businesses and middle-class earners invest in the stock market, but without accurate forecast, investors may lose money if they buy in the wrong company. Many trading apps address this issue by showing firm profits and losses. These apps use typical machine learning yet have inefficient performance measures. The proposed model used optimized probabilistic KNN method to fix the distance metric issues in pure KNN. Most test data in present KNN concentrates on centric points, but in stock prediction, all qualities are connected, hence centric points have little effect. Instead of focal points, the suggested algorithm optimizes KNN distance probabilities to focus on pro data point values. The current KNN model must calculate the distance between every data point and test point, which uses more memory and resources. The technology eliminates comparable records to avoid verification. This cuts iterations and speeds up the process. Comparing the model against benchmark classifiers shows its efficiency.