Abstract

Time series analysis is needed for big data analytics since it organizes and models time-varying data. Big data analytics uses Holt-Winters, Exponential Smoothing, and ARIMA time series analysis. One can anticipate future values using these methods. Model data patterns and find outliers. Deep long, and short-term memory networks and neural networks can improve time series forecasts. Comparing multiple time series analysis methods is essential to finding the optimal solution. When data is seasonal, ARIMA may perform better. If the trend is linear, exponential smoothing may work better. Data qualities determine the technique utilized. Big data analytics requires time series analysis for meaningful analysis and forecasts based on historical data. The technique for time series analysis is determined by Choosing the right strategy by evaluating several approaches based on data attributes.