Skip to content
DeepForgeby svx
Intermediatetime-seriesforecastingfeature-engineeringbacktestingml

Time Series & Forecasting

31 problems·~8h·4 stages

The applied forecasting workflow, from raw timestamps to monitored production forecasts. You will engineer lag and rolling features, establish baselines, fit smoothing and seasonal models, then validate with rolling-origin backtests and decay monitoring. Complements Time Series Forecasting with a stronger machine-learning and ops flavour.

Your progress

0/31 · 0%

Path readiness 0%Easy 0% · Medium 0% · Hard 0%

Continue path · Lag Operator Value

What you will learn

  • Engineer lag, rolling, and calendar features from raw series
  • Establish baselines and score forecasts with robust metrics
  • Fit exponential smoothing and seasonal models
  • Validate forecasts with backtests and monitor production decay

Before you start

1Features and Windows

0/8

Turn a series into supervised rows: the lag operator, lag-k series, rolling means and stats, windowing counts, series padding, and Fourier and time-encoding features.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.

2Baselines and Error Metrics

0/8

Never model before you have a baseline: seasonal naive forecasts, persistence, moving-average and exponential baselines, MAE, RMSE, and the seasonal MASE denominator.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.

3Smoothing and Decomposition

0/7

Separate level, trend, and season: log transforms, STL remainder values, detrending by differencing, simple and double exponential smoothing, seasonal indices, and Holt-Winters.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.

4Backtesting and Production

0/8

Trust the forecast in production: rolling-origin evaluation counts, expanding training sizes, backtest window ranges, retrain cadence, performance-decay monitoring, online updates, Theil's U, and Diebold-Mariano statistics.

Stage checkpoint

0/2 to pass

A checkpoint is a short retrieval test drawn from this stage: solve 2 of 3 from memory to mark the stage complete. The stage stays open either way.

Tried this set and did not pass? Record it — the next set swaps in fresh problems from the same stage.

Capstone

Collection

Close the path with one real artifact from the catalog.

Time Series Forecasting Arsenal

Baselines, exponential smoothing, ARMA, and GARCH as one classical toolkit.