A Study on Wordle Game data Analysis and Prediction Methods using Various Algorithms and Improvement Approaches

This article analyzes Wordle game stats, using diverse models to explain current trends and predict future outcomes. We start with Pearson Correlation Algorithm (PCCS) to find relationships between variables and Nonlinear Least Square (NLS) to explain report number changes, indicating a strong positive correlation over time. Next, Long-Short Term Memory (LSTM) predicts future reports. On March 1, 2023, we predict 59,495 reports with R-squared at 0.98949. K-means clustering and Apriori Association Rule (AAR) reveal patterns for words with two vowels. From this process, we conclude that the words with exactly two vowels have discerned tendency in numbers. Partial Least Square (PLS) models predict word gaming performance. For "EERIE" on March 1, 2023, attempts break down as follows: Try1=0.276%, Try2=4.072%, Try3=19.345%. Lastly, words classify into three difficulty levels. "EERIE" rates 2.08, as "Normal." This study enhances player understanding of the game and offers intriguing data for future research.