Time Series Forecasting

Built a model to Forecast monthly sales of Wine for certain Wine Estate for the next 12 months

Analyzed historical monthly sales data of a company. Created multiple forecast models for two different products of a particular Wine Estate and recommended the optimum forecasting model to predict monthly sales for the next 12 months along with appropriate lower and upper confidence limits.

Forecasting Models used-

1. Linear Regression
2. Naive Bayes
3. Simple Average
4. 2-pt Moving Average
5. 4-pt Moving Average
6. 6-pt Moving Average
7. 9-pt Moving Average
8. Single Exponential Smoothing
9. Double Exponential Smoothing (Holt’s Model)
10. Triple Exponential Smoothing (Holt-Winter Model)
11. ARIMA / SARIMA (Auto fitted)
12. ARIMA / SARIMA (Manually fitted)

Metric used for Comparision on models is --> RMSE on Test data

Skills and Tools

Exploratory Data Analysis on Time Series Data, Exponential Smoothing Models, ARIMA / SARIMA Models, Moving Average Models