Time Series Forecasting

intermediatedataMin 64K context

Builds and evaluates time series forecasting models using statistical and ML approaches. Supports ARIMA, Prophet, LSTM, Transformer-based models, and foundation models like TimesFM. Handles seasonality detection, trend decomposition, anomaly detection, multi-step forecasting, and backtesting with proper train/test splits for financial, IoT, and scientific time series data.

time-seriesforecastingprophetarimalstmanomaly-detectionseasonalityprediction

Use Cases

  • Demand forecasting for inventory and supply chain planning
  • Financial time series prediction with confidence intervals
  • IoT sensor anomaly detection and predictive maintenance
  • Seasonal decomposition and trend analysis
  • Multi-variate forecasting with exogenous variables
  • Backtesting and model comparison across multiple approaches

Example Prompt

I have daily sales data for the past 3 years with clear weekly and yearly seasonality.
Build a forecasting pipeline that:

1. Performs exploratory analysis (trend, seasonality, stationarity tests)
2. Decomposes the series (STL decomposition)
3. Fits multiple models:
   - SARIMA (with auto parameter selection)
   - Prophet (with holidays and regressors)
   - LightGBM with lag features
4. Evaluates using time series cross-validation (expanding window)
5. Reports MAPE, RMSE, and coverage of prediction intervals
6. Generates a 90-day forecast with 80% and 95% confidence bands
7. Detects anomalies in historical data

Output complete Python code using statsmodels, prophet, and scikit-learn.

Recommended Models

claude-opus-4-6gpt-5gemini-2-5-pro

Compatible Tools

claude-codecursorkiroany

Modalities

Input: text, code, file
→
Output: text, code

Related Skills

data-analysisscientific-visualization

Author

OpenModels Community

@openmodelsrun