A dive into time series for the energy sector
summary
Covers concrete time‑series workflows for energy: imputation strategies tied to resolution and holidays, decomposition into trend/seasonality to spot climate‑driven shifts, and lagged feature engineering with moving averages. Shows forecasting use cases (wind/solar/hydro) using weather covariates and models like RNNs, AR models, Facebook Prophet. Presents operational monitoring examples — satellite displacement measured in centimeters to flag dam risk, power‑quality anomaly detection, and clustering that isolates residential vs commercial demand changes during the pandemic.
AI-written summary · about
chapters (15)
- 0:00Time series in energy
- 3:36AI in energy transition
- 6:07What is time series data
- 9:18NLP for energy policy
- 11:29Satellite energy monitoring
- 13:15Energy price and renewables
- 15:03Forecasting energy production
- 16:25Energy consumption factors
- 17:48Clustering consumption profiles
- 19:15Data preprocessing and EDA
- 21:39Time series decomposition
- 24:05Anomaly detection in energy
- 25:49Feature engineering for models
- 26:33Forecasting models and evaluation
- 27:38Conclusion and future
related talks
40mDeep Learning and Time Series Forecasting for Smarter Energy
01h 43m(Serious) Time for Time Series
01h 30mTime Series Analysis with StatsModels
44mMethods and Tools for Time Series Data Science Problems
01h 18m(Serious) Time for Time Series
45mForecasting at Scale
32mAdopting Open-Source Tools for Time Series Forecasting
24mMachine learning and financial time series analysis