Justin Heinermann

Precise Wind Power Prediction with SVM Ensemble Regression

In this work, we propose the use of support vector regression ensembles for wind power prediction. Ensemble methods often yield better classification and regression accuracy than classical machine learning algorithms and reduce the computational cost. In the field of wind power generation, the integration into the smart grid is only possible with a precise forecast computed in a reasonable time. Our support vector regression ensemble approach uses bootstrap aggregating (bagging), which can easily be parallelized. A set of weak predictors is trained and then combined to an ensemble by aggregating the predictions. We investigate how to choose and train the individual predictors and how to weight them best in the prediction ensemble. In a comprehensive experimental analysis, we show that our SVR ensemble approach renders significantly better forecast results than state-of-the-art predictors.

Wind Power Prediction with Machine Learning

Better prediction models for the upcoming supply of renewable energy are important to decrease the need of controlling energy provided by conventional power plants. Especially for successful power grid integration of the highly volatile wind power production, a reliable forecast is crucial. In this chapter, we focus on short-term wind power prediction and employ data from the National Renewable Energy Laboratory (NREL), which are designed for a wind integration study in the western part of the United States. In contrast to physical approaches based on very complex differential equations, our model derives functional dependencies directly from the observations. Hereby, we formulate the prediction task as regression problem and test different regression techniques such as linear regression, k-nearest neighbors and support vector regression. In our experiments, we analyze predictions for individual turbines as well as entire wind parks and show that a machine learning approach yields feasible results for short-term wind power prediction.

On GPU-Based Nearest Neighbor Queries for Large-Scale Photometric Catalogs in Astronomy

Nowadays astronomical catalogs contain patterns of hundreds of millions of objects with data volumes in the terabyte range. Upcoming projects will gather such patterns for several billions of objects with peta- and exabytes of data. From a machine learning point of view, these settings often yield unsupervised, semi-supervised, or fully supervised tasks, with large training and huge test sets. Recent studies have demonstrated the effectiveness of prototype-based learning schemes such as simple nearest neighbor models. However, although being among the most computationally efficient methods for such settings (if implemented via spatial data structures), applying these models on all remaining patterns in a given catalog can easily take hours or even days. In this work, we investigate the practical effectiveness of GPU-based approaches to accelerate such nearest neighbor queries in this context. Our experiments indicate that carefully tuned implementations of spatial search structures for such multi-core devices can significantly reduce the practical runtime. This renders the resulting frameworks an important algorithmic tool for current and upcoming data analyses in astronomy.

Analysis of Diversity Methods for Evolutionary Multi-objective Ensemble Classifiers

Ensemble classifiers are strong and robust methods for classification and regression tasks. Considering the balance between runtime and classifier accuracy the learning problem becomes a multi-objective optimization problem. In this work, we propose an evolutionary multi-objective algorithm based on non-dominated sorting that balances runtime and accuracy properties of nearest neighbor classifier ensembles and decision tree ensembles. We identify relevant ensemble parameters with a significant impact on the accuracy and runtime. In the experimental part of this paper, we analyze the behavior on typical classification benchmark problems.

Short-Term Wind Power Prediction with Combination of Speed and Power Time Series

The integration of wind power generation into the power grid can only succeed with precise and reliable forecast methods. With different measurements available, machine learning algorithms can yield very good predictions for short-term forecast horizons. In this paper, we compare the use of wind power and wind speed time series as well as differences of subsequent measurements with Random Forests, Support Vector Regression and k-nearest neighbors. While both time series, wind power and speed, are well-suited to train a predictor, the best performance can be achieved by using both together. Further, we propose an ensemble approach combining RF and SVR with a cross-validated weighted average and show that the prediction performance can be substantially improved.

A framework for data mining in wind power time series

Wind energy is playing an increasingly important part for ecologically friendly power supply. The fast growing infrastructure of wind turbines can be seen as large sensor system that screens the wind energy at a high temporal and spatial resolution. The resulting databases consist of huge amounts of wind energy time series data that can be used for prediction, controlling, and planning purposes. In this work, we describe WindML, a Python-based framework for wind energy related machine learning approaches. The main objective of WindML is the continuous development of tools that address important challenges induced by the growing wind energy information infrastructures. Various examples that demonstrate typical use cases are introduced and related research questions are discussed. The different modules of WindML reach from standard machine learning algorithms to advanced techniques for handling missing data and monitoring high-dimensional time series.

Preferences-Based Choice Prediction in Evolutionary Multi-objective Optimization

Evolutionary multi-objective algorithms (EMOAs) of the type of NSGA-2 approximate the Pareto-front, after which a decision-maker (DM) is confounded with the primary task of selecting the best solution amongst all the equally good solutions on the Pareto-front. In this paper, we complement the popular NSGA-2 EMOA by posteriori identifying a DM’s best solution among the candidate solutions on the Pareto-front, generated through NSGA-2. To this end, we employ a preference-based learning approach to learn an abstract ideal reference point of the DM on the multi-objective space, which reflects the compromises the DM makes against a set of conflicting objectives. The solution that is closest to this reference-point is then predicted as the DM’s best solution. The pairwise comparisons of the candidate solutions provides the training information for our learning model. The experimental results on ZDT1 dataset shows that the proposed approach is not only intuitive, but also easy to apply, and robust to inconsistencies in the DM’s preference statements.

Evolutionary Multi-objective Ensembles for Wind Power Prediction

Ensembles turn out to be excellent wind power prediction methods. But the space of algorithms and parameters of supervised learning ensembles is large. For an efficient optimization and tuning of ensembles, we propose to employ evolutionary multi-objective optimization methods in this work. NSGA-II is a classic optimization algorithm based on non-dominated sorting and maximization of the crowding distance and has successfully been applied in various applications in the past. The experimental part of the paper shows how NSGA-II tunes SVR ensembles, random forests, and heterogenous ensembles. The study demonstrates that the proposed approach evolves an attractive set of ensembles for a practitioner yielding numerous compromises of prediction accuracy and runtime.