Global Energy Forecasting Competition (GEFCom 2014)
GEFCom 2014 (proenfo_gfc14) provides 17,520 hourly observations across a single electrical zone, targeting a 168-hour (7-day ahead) forecast of power grid demand. Conditioned on exogenous air temperature, the benchmark tests medium-term multi-seasonal grid scheduling and dispatch optimization.
targetfev / AutoGluon ↗Business Objective: Global Energy Forecasting Competition (GEFCom 2014)
01Business Context
Transmission system operators (TSOs) and energy trading desks operate on weekly commitment cycles, scheduling baseload and peaker generation 168 hours ahead to ensure grid stability and clear day-ahead power markets.
02Analytical Objective
Forecast hourly zonal electrical load across a full 168-hour ahead horizon using historical consumption telemetry and known future ambient air temperatures.
03Economic & Decision Impact
Under-forecasting risks reserve margin depletion, localized brownouts, and costly spot-market imbalance penalties. Over-forecasting forces unnecessary thermal plant spinning reserves, curtailment of renewable generation, and elevated operating expenditures.
04Temporal Dynamics & Covariate Information Flow
Dominant Seasonality: Strong 24-hour diurnal cycle combined with a 168-hour weekly profile showing distinct weekend baseload drops, overlaid with non-linear heating/cooling seasonal shifts.
Covariates & Information Flow: Future surface air temperature forecasts (`airtemperature`) directly dictate thermo-dependent electric heating and HVAC cooling loads, alongside calendar regressors.
Classical univariate baselines (AutoARIMA, ETS) struggle over the long 168-hour horizon due to phase error accumulation across multi-day diurnal cycles. Feature-engineered LightGBM models with temperature lag interactions excel, while zero-shot foundation models such as Mistral's Timestral and MOIRAI capture complex non-linear load-temperature dynamics without manual feature crafting.
Forecast this dataset in your browser
A representative series is read directly from this dataset’s Parquet file. Move the forecast start, compare held-out observations, and toggle available numeric covariates with Timestral 5M (12 MB) or switch to ci-proto (~346 MB, WebGPU). Benchmark scores above use the stated evaluation setup.
Target & Feature Column Definitions
Exhaustive business definitions, measurement units, roles, and target variables across all 4 columns.
Known Future Covariates vs. Historically Observed Covariates
Multi-step forecasting models must respect the arrow of time and real-world information availability. FEV benchmarks and Carla strictly distinguish between features known across the forward prediction window versus realizations available only up to the forecast origin.
Scheduled promotions (such as Rossmann's Promo), calendar events (school/state holidays), and operational forecasts (such as Nord Pool's day-ahead grid load and wind power forecasts). These are deterministic or published before market clearing, allowing foundation models to legitimately condition on them across future steps.
Stochastic operational observations (such as Rossmann's actual customer footfall Customers, Rohlik's actual precipitation and snowfall, or sudden warehouse shutdowns). Because these are only realized in real-time, future values must never enter a prediction to prevent lookahead target leakage.
| Col | Variable / Feature Name | Role | Type | Unit / Scale | Description & Business Meaning |
|---|---|---|---|---|---|
| C | targetZonal Electrical Load | TARGET | NUM | MW | Total hourly electric power consumption observed across the regional utility zone. |
| A | idSeries Identifier | SERIES ID | CAT | — | Unique categorical identifier denoting the electrical zonal load time series. |
| B | timestampTimestamp | TIME | TIME | — | Hourly observation date and time in ISO 8601 format. |
| D | airtemperatureAir Temperature | FUTURE KNOWN | NUM | deg F | Observed and forecasted ambient surface air temperature influencing thermal building demand. • Known future driver: Legitimate lookahead across forecast horizon (steps t+1 to t+H, e.g. calendar/promotion/forecast). |
Interactive Data Table
Explore rows, feature values, and target labels for Global Energy Forecasting Competition (GEFCom 2014).
Foundation Models vs Traditional ML & Baselines (35 Methods)
Empirical evaluation directly from the AutoGluon fev benchmark suite. Global aggregate ranking: The win rates and skill scores below represent cross-dataset performance across all 100 standardized tasks in fev-bench, rather than results isolated to Global Energy Forecasting Competition (GEFCom 2014).
| Rank | Method / Model | Family | Organization | SQL Win Rate?Average win rate (%) against all competing models across the benchmark tasks under Scaled Quantile Loss. | SQL Skill?Normalized skill score relative to the Seasonal Naive baseline (0% = baseline, higher is better). | MASE Win Rate?Average win rate (%) under Mean Absolute Scaled Error. | FEV Latency?FEV benchmark reference latency measured in server Python (median seconds per 100 series). Independent in-browser WebGPU execution for Carla is reported in the telemetry card above. | Link |
|---|---|---|---|---|---|---|---|---|
| #1 | ci-proto | Foundation Model | Continuous Intelligence | 90.0% | +51.5% | 88.4% | 0.53s | Repo ↗ |
| #2 | TimeH2O-v1 | Foundation Model | H2O.ai | 89.7% | +50.8% | 88.2% | 2.82s | Repo ↗ |
| #3 | Timestral-209M | Foundation Model | Mistral AI | 87.1% | +50.4% | 85.4% | 0.51s | Repo ↗ |
| #4 | t0-gamma | Foundation Model | The Forecasting Company | 84.7% | +49.4% | 82.0% | 4.38s | Repo ↗ |
| #5 | Simons-297M | Foundation Model | Abel AI Lab | 82.0% | +48.4% | 80.9% | 0.94s | Repo ↗ |
| #6 | timesfm-3 | Foundation Model | 79.3% | +49.0% | 77.6% | 3.41s | Repo ↗ | |
| #7 | iAmTime-base-forecast | Foundation Model | Walmart Labs | 73.7% | +46.8% | 72.5% | 0.56s | Repo ↗ |
| #8 | Chronos-2 | Foundation Model | AWS | 72.2% | +47.3% | 69.9% | 0.84s | Repo ↗ |
| #9 | t0-beta | Foundation Model | The Forecasting Company | 69.9% | +46.6% | 66.9% | 2.29s | Repo ↗ |
| #10 | Toto-2.0-2.5B | Baseline | Community | 69.4% | +44.4% | 67.8% | 5.58s | Repo ↗ |
| #11 | TiRex-2 | Foundation Model | NX-AI | 69.2% | +45.5% | 66.3% | 0.27s | Repo ↗ |
| #12 | TiRex | Foundation Model | NX-AI | 62.2% | +42.6% | 58.0% | 0.24s | Repo ↗ |
| #13 | Toto-2.0-22m | Baseline | Community | 61.6% | +42.9% | 59.4% | 0.25s | Repo ↗ |
| #14 | TimesFM-2.5 | Baseline | Community | 58.4% | +42.2% | 56.7% | 1.87s | Repo ↗ |
| #15 | TabPFN-TS-3 | Foundation Model | Prior Labs | 55.5% | +43.1% | 53.3% | 235s | Repo ↗ |
| #16 | TS-ICL | Foundation Model | EDF-Lab | 55.1% | +43.1% | 53.4% | 2.49s | Repo ↗ |
| #17 | citras-fm | Foundation Model | Hitachi | 52.9% | +41.2% | 50.6% | 0.60s | Repo ↗ |
| #18 | FlowState | Foundation Model | IBM | 52.7% | +39.9% | 53.0% | 2.29s | Repo ↗ |
| #19 | Toto-1.0 | Baseline | Community | 51.6% | +40.8% | 48.8% | 22.35s | Repo ↗ |
| #20 | TabPFN-TS | Foundation Model | Prior Labs | 50.5% | +41.5% | 48.0% | 109s | Repo ↗ |
| #21 | Moirai-2.0 | Baseline | Community | 46.2% | +39.3% | 44.2% | 0.35s | Repo ↗ |
| #22 | Chronos-Bolt | Foundation Model | AWS | 45.4% | +38.9% | 42.9% | 0.26s | Repo ↗ |
| #23 | TFT | Deep Learning | — | 34.7% | +32.2% | 31.6% | 829s | Repo ↗ |
| #24 | Sundial-Base | Foundation Model | Tsinghua University | 30.8% | +33.4% | 37.0% | 8.01s | Repo ↗ |
| #25 | Stat. Ensemble | Baseline | Community | 30.6% | +20.7% | 33.6% | 147s | Repo ↗ |
| #26 | PatchTST | Deep Learning | — | 30.0% | +30.1% | 27.8% | 664s | Repo ↗ |
| #27 | DeepAR | Deep Learning | — | 28.9% | +29.1% | 25.9% | 1113s | Repo ↗ |
| #28 | AutoARIMA | Statistical Baseline | — | 26.6% | +21.3% | 24.2% | 20.14s | Repo ↗ |
| #29 | AutoETS | Statistical Baseline | — | 22.5% | -26.8% | 24.1% | 3.48s | Repo ↗ |
| #30 | CatBoost | Tree Ensemble | — | 21.8% | +23.0% | 37.5% | 29.48s | Repo ↗ |
| #31 | LightGBM | Tree Ensemble | — | 20.3% | +21.0% | 35.7% | 3.03s | Repo ↗ |
| #32 | AutoTheta | Statistical Baseline | — | 17.9% | +5.5% | 22.2% | 3.28s | Repo ↗ |
| #33 | Seasonal Naive | Baseline | Community | 12.0% | 0.0% | 13.2% | 0.48s | Repo ↗ |
| #34 | Naive | Statistical Baseline | — | 7.8% | -45.4% | 12.2% | 0.47s | Repo ↗ |
| #35 | Drift | Statistical Baseline | — | 7.0% | -45.8% | 11.1% | 0.45s | Repo ↗ |
Variable Schema & Summary Distributions
Observed numerical ranges, category cardinalities, missing rates, and sample values across 17,520 rows.
| Col | Variable Name | Type | Missing | Distinct | Summary Stats / Distribution | Sample Values |
|---|---|---|---|---|---|---|
| C | targetTARGET | NUM | 0 | 2,502 | [2107, 5224] μ=3328.2 σ=561.3 | 286627102605 |
| A | id | CAT | 0 | 1 | — | loadloadload |
| B | timestamp | CAT | 0 | 17,520 | — | 2013-01-01T01:00:002013-01-01T02:00:002013-01-01T03:00:00 |
| D | airtemperatureFUTURE KNOWN | NUM | 0 | 311 | [-13.670000076293945, 95] μ=46.8 σ=19.8 | 29.32999992370605529.32999992370605530 |
Frequently Asked Questions: Global Energy Forecasting Competition (GEFCom 2014)
Common questions regarding the Global Energy Forecasting Competition (GEFCom 2014) dataset, machine learning task formulations, and in-browser tabular inference.
What is the forecasting horizon for GEFCom 2014 (proenfo_gfc14)?
The horizon is 168 hours (7 full calendar days), reflecting standard transmission grid scheduling and weekly generation dispatch cycles.
Can I run zero-server forecasts on this dataset in Google Sheets?
Yes, Carla enables zero-server in-context time series forecasting directly in spreadsheets using Mistral's Timestral foundation model without moving data outside your sheet.
Which models perform best on GEFCom 2014?
Modern zero-shot time series foundation models like Mistral's Timestral and tuned gradient-boosted trees (LightGBM) with temperature interaction terms outperform classical statistical baselines like AutoARIMA and Exponential Smoothing.
Forecast with Carla: Global Energy Forecasting Competition (GEFCom 2014)
Open the pre-loaded Google Sheet and let Carla generate multi-step probabilistic forecasts using Mistral's Timestral directly inside your browser with zero server uploads.
Launch Carla
Click once to open the spreadsheet and Carla side panel together.
Review the Setup
Carla auto-detects the target sequence, frequency cadence, and temporal covariates.
Forecast & Calibrate
Generate multi-step quantile forecasts with zero server uploads.
Provenance & Attribution
Hong, T., Pinson, P., Fan, S., Zareipour, H., Troccoli, A., & Hyndman, R. J. (2016). Probabilistic energy forecasting: Global Energy Forecasting Competition 2014 and beyond. International Journal of Forecasting, 32(3), 896-913.