Erik Pitzer

• DI (FH) Dr.
• Professor at University of Applied Sciences Upper Austria

As the digital transformation of industry continues, more and more data is being collected to gain insights into and further improve existing processes, known as prescriptive analytics. Among the enabling technologies for prescriptive analytics is simulation-based optimization. To accelerate the execution of simulations, the approach can be coupled...

Dynamic optimization problems pose a big challenge for classic optimization algorithms. They could simply be viewed as a series of related optimization problems. In particular the aspect of time-linkage has not been well studied yet. In this work we are analyzing an artificial problem based on real-world data to elucidate the potential of fitness l...

Dynamic optimization problems are of significant practical relevance, but suffer from a lack of analysis. The characteristics of time-linked problems are especially difficult to capture as future problem states depend on the optimizer’s performance. By tracking numeric features along the optimization process, information about the problems characte...

With the growing use of machine learning models in many critical domains, research regarding making these models, as well as their predictions, more explainable has intensified in the last few years. In this paper, we present extensions to the machine learning based data mining technique Variable Interaction Networks (VIN), to integrate existing do...

The typical methods for symbolic regression produce rather abrupt changes in solution candidates. In this work, we have tried to transform symbolic regression from an optimization problem, with a landscape that is so rugged that typical analysis methods do not produce meaningful results, to one that can be compared to typical and very smooth real-v...

Incremental evaluation is a big advantage for trajectory-based optimization algorithms. Previously, the application of similar ideas to crossover-based algorithms, such as genetic algorithms did not seem appealing as the expected benefit would be marginal. We propose the use of an immutable data structure that stores partial evaluation results insi...

Exploratory fitness landscape analysis (FLA) is a category of techniques that try to capture knowledge about a black-box optimization problem. This is achieved by assigning features to a certain problem instance utilizing only information obtained by evaluating the black-box. This knowledge can be used to obtain new domain knowledge but more often...

The main idea of this paper is to add model set pre-processing for Genetic Programming based Evolvement of Models of Models. Simple Symbolic Formulas generated offline with the help of the deterministic function extraction algorithm will be used as building blocks for Genetic Programming. In this work, a pre-processing of models set is generated by...

Exploratory landscape analysis is a useful method for algorithm selection, parametrization and creating an understanding of how a heuristic optimization algorithm performs on a problem and why. A prominent family of fitness landscape analysis measures are based on random walks through the search space. However, most of these features were only intr...

Evolutionary algorithm analysis is often impeded by the large amounts of intermediate data that is usually discarded and has to be painstakingly reconstructed for real-world large-scale applications. In the recent past persistent data structures have been developed which offer extremely compact storage with acceptable runtime penalties. In this wor...

Much of the literature found on surrogate models presents new approaches or algorithms trying to solve black-box optimization problems with as few evaluations as possible. The comparisons of these new ideas with other algorithms are often very limited and constrained to non-surrogate algorithms or algorithms following very similar ideas as the pres...

The no free lunch (NFL) theorem puts a limit to the range of problems a certain metaheuristic algorithm can be applied to successfully. For many methods these limits are unknown a priori and have to be discovered by experimentation. With the use of fitness landscape analysis (FLA) it is possible to obtain characteristic data and understand why meth...

A major drawback of surrogate-assisted evolutionary algorithms is their limited ability to perform in high-dimensional scenarios. This paper describes a possible meta-algorithm scheme for the application of surrogate models to high-dimensional optimization problems. The main assumption of the proposed method is that for some of these expensive prob...

The typical methods for symbolic regression produce rather abrupt changes in solution candidates. In this work, we have tried to transform symbolic regression from an optimization problem, with a landscape that is so rugged that typical analysis methods do not produce meaningful results, to one that can be compared to typical and very smooth real-v...

Dynamic and stochastic problem environments are often difficult to model using standard problem formulations and algorithms. One way to model and then solve them is simulation-based optimization: Simulations are integrated into the optimization process in order to evaluate the quality of solution candidates and to identify optimized system configur...

Optimization of supply chains and logistic networks have been largely addressed by simulation-based optimization. The logistics for low-energy biological residues poses a great challenge for logistics that has to be tackled at an international scale. For this application, a new simplified model of logistic networks was created that allows not only...

The quadratic assignment problem is among the harder combinatorial optimization problems in that even small instances might be difficult to solve and for different algorithms different instances pose challenges to solve. Research on the quadratic assignment problem has thus focused on developing methods that defy the problem’s variety and that can...

Estimating hardness based on intrinsic characteristics of problem instances plays an important role in algorithm selection and parameter tuning. We have compiled an extensive study of different fitness landscape and problem specific measures for the quadratic assignment problem to predict and correlate problem instance hardness for several differen...

Many optimization problems cannot be solved by classical mathematical optimization techniques due to their complexity and the size of the solution space. In order to achieve solutions of high quality though, heuristic optimization algorithms are frequently used. These algorithms do not claim to find global optimal solutions, but offer a reasonable...

In the last few years, fitness landscape analysis has seen an increase in interest due to the availability of large problem collections and research groups focusing on the development of a wide array of different optimization algorithms for diverse tasks. Instead of being able to rely on a single trusted method that is tuned and tweaked to the appl...

Metaheuristic optimization algorithms are general optimization strategies suited to solve a range of real-world relevant optimization problems. Many metaheuristics expose parameters that allow to tune the effort that these algorithms are allowed to make and also the strategy and search behavior [1]. Adjusting these parameters allows to increase the...

In this work we elaborate on the measurement of anisotropy in fitness landscapes by defining an extension over arbitrary base measures. This rather pragmatic method’s soundness is justified by statistical argument and tested on several existing and new fitness landscapes. Moreover, new variants of the popular NK landscapes are introduced that exhib...

In this paper we investigate how to stack products on a storage yard for efficient retrieval. The objective is to minimise both the transport distance and the number of stack shuffles. Previous research on yard storage assignment indicated that the fitness landscape of the problem features a high degree of neutrality, meaning that there are many ne...

This paper describes the optimization knowledge base (OKB), a database for storing information about algorithms and problems. The optimization knowledge base allows to save results of algorithm executions as well as problem-specific information of fitness landscape analyses. This database can be queried and gives researchers a tool for gaining a be...

Fitness landscape analysis methods have become an increasingly popular topic for research. The future application of these methods to metaheuristics can yield advanced self-adaptive metaheuristics and knowledge bases that can take the role of expert systems in the field of optimization. One important feature of such an expert system would be the pr...

Optimization of simulation parameters is an important task in many different sciences where simulation is used to model and analyze complex processes and behaviors. In this work it is shown how users, such as researchers, students, and practitioners can benefit from the integration of data-exchange-interfaces in optimization software system. The de...

Metaheuristics are successfully applied in many different application domains as they provide a reasonable tradeoff between computation time and achievable solution quality. However, choosing an appropriate algorithm for a certain problem is not trivial, as problem characteristics can change remarkably for different instances and the performance of...

In this paper the fitness landscape of a simulation optimisation problem is analysed within the metaheuristic optimisation framework Heuristic Lab. Computational experiments are performed within an application prototype of a link between the model of a vehicle scheduling problem and the optimisation framework. Modern fitness landscape analysis appr...

In the past, the notion of fitness landscapes has found widespread adoption. Many different methods have been developed that provide a general and abstract framework applicable to any optimization problem. We formally define fitness landscapes, provide an in-depth look at basic properties and give detailed explanations and examples of existing fitn...

Solutions to the Quadratic Assignment Problem (QAP) can be compared with each other in several ways. In this work a new distance metric for measuring the distance respectively similarity between two solutions will be introduced. Such a metric is useful in measuring the performance of heuristic optimization algorithms and generally in the analysis o...

Production Fine Planning is often performed directly using all information and assuming that it is fixed. In practice, however, this information changes regularly and the plan has to be adapted. This often means a complete rescheduling of all operations. We present a new approach to this problem by optimizing priority rules that can sort the availa...

Many different techniques have been developed for fitness landscape analysis. We present a consolidated and uniform implementation inside the heuristic optimization platform HeuristicLab. On top of these analysis methods a new approach to empirical measurement of isotropy is presented that can be used to extend existing methods. Results are shown u...

Formal fitness landscape analysis enables us to study basins of attraction with great detail. This seemingly simple concept shows more variability and influence on heuristic algorithms than might be expected. We have taken a new perspective and a closer look at the properties of basins of attraction using two-dimensional visualizations to arrive at...

The Gene Expression Omnibus (GEO) is the largest resource of public gene expression data. While GEO enables data browsing, query and retrieval, additional tools can help realize its potential for aggregating and comparing data across multiple studies and platforms. This paper describes DSGeo-a collection of valuable tools that were developed for an...

Microarray probes and reads from massively parallel sequencing technologies are two most widely used genomic tags for a transcriptome study. Names and underlying technologies might differ, but expression technologies share a common objective-to obtain mRNA abundance values at the gene level, with high sensitivity and specificity. However, the initi...

Large repositories of biomedical research data are most useful to translational researchers if their data can be aggregated for efficient queries and analyses. However, inconsistent or non-existent annotations describing important sample details such as name of tissue or cell line, histopathological type, and subject characteristics like demographi...

This study describes a large-scale manual re-annotation of data samples in the Gene Expression Omnibus (GEO), using variables and values derived from the National Cancer Institute thesaurus. A framework is described for creating an annotation scheme for various diseases that is flexible, comprehensive, and scalable. The annotation structure is eval...

De novo peptide sequencing algorithms are often tested on relatively small data sets made of excellent spectra. Since there are always more and more tandem mass spectra available, we have assembled six large, reliable, and diverse (three mass spectrometer types) data sets intended for such tests and we make them accessible via a web server. To exem...

Plugin-based software systems are the next step of evolution in application development. By supporting fine grained modularity not only on the source code but also on the post-compilation level, plugin frameworks help to handle complexity, simplify application configuration and deployment, and enable users or third parties to easily enhance existin...

Multiple Sequence Alignment is an essential tool in the analysis and comparison of biological sequences. Unfortunately, the complexity of this problem is exponential. Currently feasible methods are, therefore, only approximations. The progressive multiple sequence alignment algorithms are the most widespread among these approximations. Still, the c...