Prompt

Develop comprehensive time series forecasts tailored for business planning by following these structured steps:

### 1. Define the Forecasting Objective
- **Objective**: Clearly state the purpose of the forecast (e.g., sales prediction, inventory management).

### 2. Choose Appropriate Forecasting Methods
Select one or more of the following methods based on your data characteristics:
- **Trend Analysis**: Identify long-term movement in data.
- **Seasonality Detection**: Analyze periodic fluctuations.
- **ARIMA Modeling**: Use for non-stationary data with trends and seasonality.
- **Exponential Smoothing**: Apply for data with trends and seasonal patterns.
- **Machine Learning Approaches**: Consider advanced techniques like LSTM or Random Forest for complex datasets.

### 3. Model Selection Criteria
Evaluate models based on:
- **Data Characteristics Assessment**: Understand the nature of your data (e.g., frequency, seasonality).
- **Accuracy Metrics Comparison**: Use metrics like MAE, RMSE, or MAPE for evaluation.
- **Cross-Validation Strategies**: Implement techniques like k-fold or time-series split.
- **Ensemble Techniques**: Combine multiple models for improved accuracy.
- **Uncertainty Quantification**: Assess the confidence intervals of forecasts.

### 4. Integrate with Business Processes
Ensure your forecasts align with business needs by:
- **Scenario Planning**: Develop different scenarios based on forecast outcomes.
- **Risk Assessment**: Identify potential risks associated with forecasts.
- **Resource Allocation**: Plan resources based on forecasted demand.
- **Performance Monitoring**: Track forecast accuracy over time.
- **Forecast Updating**: Regularly revise forecasts with new data.

### 5. Input Requirements
- **Time Series Data**: {INSERT DATA TYPE} (e.g., daily sales data, monthly web traffic).
- **Forecast Horizon**: {INSERT HORIZON} (e.g., next 3 months, next fiscal year).

### 6. Edge Cases
- Address potential issues such as missing data, outliers, or sudden market changes. Specify how to handle these scenarios (e.g., imputation methods, anomaly detection).

### Success Criteria
- Define what success looks like (e.g., achieving a forecast accuracy of less than 10% error).