Data services and Training
We provide analytical solutions to challenging problems in the energy industry. In particular, we help consumers and utilities to make better decisions by answering some questions like:
- Which is the impact of the adoption of smart grid technologies, such as electric vehicles, storage technologies, smart meters, etc., on the electricity demand profile?
- How an increase of renewable distributed generation will affect the operation and planning of distribution networks?
- Which type of consumers will increase their risk as a function of their electricity or gas demand profiles?
- How to derive optimal real time pricing strategies for electricity and gas consumers?
- How to obtain accurate predictions for spot and futures market prices for both electricity and demand?
- Which is the optimal selling and buying strategy to succesfully hedge risk in electricity and gas markets?
- How to efficiently manage renewable generation deviations and reserves?
To answer these questions, we provide training and consulting services. Some of the training modules we offer include:
Basic tools for data analysis in R / Python
- Exploratory analysis
- Data visualization
- Inference: prediction error, confidence intervals and hypothesis testing
Supervised learning: Classification techniques
- Logistic Regression
- Bayes Classifiers
- Naive Bayes
- Nearest Neighbors
- SVMs
- Boosting, Bagging
- Decision Trees and Random Forests
Time series and forecasting
- ARIMAs
- Exponential Smoothing
- Dynamic Regression
- Multivariate Time Series
- Combination of Forecasts
- Volatily and Risk Forecasting
Decision Making Under Uncertainty
- Risk Management
- Simulation techniques
- Stochastic Programming
- Worst-case Analysis
- Robust Optimization
Supervised learning: Regression Models and Prediction
- Simple Regression
- Multiple Regression
- Advanced Tools: Nonlinear models, Model Selection, Ridge and Lasso Regression
Unsupervised Learning
- PCA and SVD
- Shrinkage Estimation
- Factor Analysis
- Clustering
- Non-hierarchical Methods, Hierarchical Methods
Optimization
- Unconstrained Models
- Linear Models
- Discrete Models and Networks
- Nonlinear Models
- Complementarity Modeling
- Constraint Learning
Resampling techniques
- Cross-validation
- Bootstrap
- Jackknife
- Dynamic data