samkamau81/Kenyas-Maize-Output-Time-Series

This is a time series project that utilizes Facebook prophet to predict what Kenya's Maize Output will be in 2027 # Kenyas-Maize-Output-Time-Series This is a time series project that utilizes Facebook prophet to predict what Kenya's Maize Output will be in 2027. Time forecasting models use historical data to predict future values of these indicators, allowing analysts to anticipate economic trends and adjust policies accordingly. I've created a simple Dashboard using Microsoft Power BI , a powerful tool for Data Visualization, to show how Kenya Maize Yield behaves interms of Imports and Exports , Location where it is grown and When is the yield mostly gotten. I have used data from FAO (Food and Agriculture Organization) to come up with the dashboard and Time series analysis. - - - - # Data Preparation ## Handling Missing Data The most common methods to address missing data in time series are: - **Imputation** When we fill in missing data based on observations about the entire data set. - **Interpolation** When we use neighboring data points to estimate the missing value. Interpolation can also be a form of imputation. - **Deletion of affected time periods** When we choose not to use time periods that have missing data at all - **Smoothing Data** __Smoothing data__ can be done for a variety of reasons, and often real-world time series data is smoothed before analysis, especially for visualizations that aim to tell an understandable story about the data. While outlier detection is a topic in and of itself, if you have reason to believe your data should be smoothed, you can do so with a moving average to eliminate measurement spikes, errors of measurement, or both. Even if the spikes are accurate, they may not reflect the underlying process and may be more a matter of instrumentation problems; this is why it’s quite common to smooth data. __Smoothing data__ is strongly related to imputing missing data, and so some of those techniques are relevant here as well. For example, you can smooth data by applying a rolling mean, with or without a lookahead, as that is simply a matt …