Refine
H-BRS Bibliography
- yes (31)
Departments, institutes and facilities
- Fachbereich Informatik (31) (remove)
Document Type
- Conference Object (31) (remove)
Year of publication
Keywords
- Quality diversity (3)
- Aerodynamics (2)
- Evolutionary computation (2)
- Heart Rate Prediction (2)
- MAP-Elites (2)
- Neuroevolution (2)
- Quality Diversity (2)
- 3D design (1)
- Autoencoder (1)
- Bayesian optimization (1)
We consider multi-solution optimization and generative models for the generation of diverse artifacts and the discovery of novel solutions. In cases where the domain's factors of variation are unknown or too complex to encode manually, generative models can provide a learned latent space to approximate these factors. When used as a search space, however, the range and diversity of possible outputs are limited to the expressivity and generative capabilities of the learned model. We compare the output diversity of a quality diversity evolutionary search performed in two different search spaces: 1) a predefined parameterized space and 2) the latent space of a variational autoencoder model. We find that the search on an explicit parametric encoding creates more diverse artifact sets than searching the latent space. A learned model is better at interpolating between known data points than at extrapolating or expanding towards unseen examples. We recommend using a generative model's latent space primarily to measure similarity between artifacts rather than for search and generation. Whenever a parametric encoding is obtainable, it should be preferred over a learned representation as it produces a higher diversity of solutions.
Neuroevolution methods evolve the weights of a neural network, and in some cases the topology, but little work has been done to analyze the effect of evolving the activation functions of individual nodes on network size, an important factor when training networks with a small number of samples. In this work we extend the neuroevolution algorithm NEAT to evolve the activation function of neurons in addition to the topology and weights of the network. The size and performance of networks produced using NEAT with uniform activation in all nodes, or homogenous networks, is compared to networks which contain a mixture of activation functions, or heterogenous networks. For a number of regression and classification benchmarks it is shown that, (1) qualitatively different activation functions lead to different results in homogeneous networks, (2) the heterogeneous version of NEAT is able to select well performing activation functions, (3) the produced heterogeneous networks are significantly smaller than homogeneous networks.
Computers can help us to trigger our intuition about how to solve a problem. But how does a computer take into account what a user wants and update these triggers? User preferences are hard to model as they are by nature vague, depend on the user’s background and are not always deterministic, changing depending on the context and process under which they were established. We pose that the process of preference discovery should be the object of interest in computer aided design or ideation. The process should be transparent, informative, interactive and intuitive. We formulate Hyper-Pref, a cyclic co-creative process between human and computer, which triggers the user’s intuition about what is possible and is updated according to what the user wants based on their decisions. We combine quality diversity algorithms, a divergent optimization method that can produce many, diverse solutions, with variational autoencoders to both model that diversity as well as the user’s preferences, discovering the preference hypervolume within large search spaces.
The initial phase in real world engineering optimization and design is a process of discovery in which not all requirements can be made in advance, or are hard to formalize. Quality diversity algorithms, which produce a variety of high performing solutions, provide a unique chance to support engineers and designers in the search for what is possible and high performing. In this work we begin to answer the question how a user can interact with quality diversity and turn it into an interactive innovation aid. By modeling a user's selection it can be determined whether the optimization is drifting away from the user's preferences. The optimization is then constrained by adding a penalty to the objective function. We present an interactive quality diversity algorithm that can take into account the user's selection. The approach is evaluated in a new multimodal optimization benchmark that allows various optimization tasks to be performed. The user selection drift of the approach is compared to a state of the art alternative on both a planning and a neuroevolution control task, thereby showing its limits and possibilities.
With the increasing average age of the population in many developed countries, afflictions like cardiovascular diseases have also increased. Exercising has a proven therapeutic effect on the cardiovascular system and can counteract this development. To avoid overstrain, determining an optimal training dose is crucial. In previous research, heart rate has been shown to be a good measure for cardiovascular behavior. Hence, prediction of the heart rate from work load information is an essential part in models used for training control. Most heart-rate-based models are described in the context of specific scenarios, and have been evaluated on unique datasets only. In this paper, we conduct a joint evaluation of existing approaches to model the cardiovascular system under a certain strain, and compare their predictive performance. For this purpose, we investigated some analytical models as well as some machine learning approaches in two scenarios: prediction over a certain time horizon into the future, and estimation of the relation between work load and heart rate over a whole training session.
Surrogate-assistance approaches have long been used in computationally expensive domains to improve the data-efficiency of optimization algorithms. Neuroevolution, however, has so far resisted the application of these techniques because it requires the surrogate model to make fitness predictions based on variable topologies, instead of a vector of parameters. Our main insight is that we can sidestep this problem by using kernel-based surrogate models, which require only the definition of a distance measure between individuals. Our second insight is that the well-established Neuroevolution of Augmenting Topologies (NEAT) algorithm provides a computationally efficient distance measure between dissimilar networks in the form of "compatibility distance", initially designed to maintain topological diversity. Combining these two ideas, we introduce a surrogate-assisted neuroevolution algorithm that combines NEAT and a surrogate model built using a compatibility distance kernel. We demonstrate the data-efficiency of this new algorithm on the low dimensional cart-pole swing-up problem, as well as the higher dimensional half-cheetah running task. In both tasks the surrogate-assisted variant achieves the same or better results with several times fewer function evaluations as the original NEAT.
A new method for design space exploration and optimization, Surrogate-Assisted Illumination (SAIL), is presented. Inspired by robotics techniques designed to produce diverse repertoires of behaviors for use in damage recovery, SAIL produces diverse designs that vary according to features specified by the designer. By producing high-performing designs with varied combinations of user-defined features a map of the design space is created. This map illuminates the relationship between the chosen features and performance, and can aid designers in identifying promising design concepts. SAIL is designed for use with compu-tationally expensive design problems, such as fluid or structural dynamics, and integrates approximative models and intelligent sampling of the objective function to minimize the number of function evaluations required. On a 2D airfoil optimization problem SAIL is shown to produce hundreds of diverse designs which perform competitively with those found by state-of-the-art black box optimization. Its capabilities are further illustrated in a more expensive 3D aerodynamic optimization task.
The MAP-Elites algorithm produces a set of high-performing solutions that vary according to features defined by the user. This technique to 'illuminate' the problem space through the lens of chosen features has the potential to be a powerful tool for exploring design spaces, but is limited by the need for numerous evaluations. The Surrogate-Assisted Illumination (SAIL) algorithm, introduced here, integrates approximative models and intelligent sampling of the objective function to minimize the number of evaluations required by MAP-Elites.
The ability of SAIL to efficiently produce both accurate models and diverse high-performing solutions is illustrated on a 2D airfoil design problem. The search space is divided into bins, each holding a design with a different combination of features. In each bin SAIL produces a better performing solution than MAP-Elites, and requires several orders of magnitude fewer evaluations. The CMA-ES algorithm was used to produce an optimal design in each bin: with the same number of evaluations required by CMA-ES to find a near-optimal solution in a single bin, SAIL finds solutions of similar quality in every bin.
The encoding of solutions in black-box optimization is a delicate, handcrafted balance between expressiveness and domain knowledge between exploring a wide variety of solutions, and ensuring that those solutions are useful. Our main insight is that this process can be automated by generating a dataset of high-performing solutions with a quality diversity algorithm (here, MAP-Elites), then learning a representation with a generative model (here, a Varia-tional Autoencoder) from that dataset. Our second insight is that this representation can be used to scale quality diversity optimization to higher dimensions-but only if we carefully mix solutions generated with the learned representation and those generated with traditional variation operators. We demonstrate these capabilities by learning an low-dimensional encoding for the inverse kinemat-ics of a thousand joint planar arm. The results show that learned representations make it possible to solve high-dimensional problems with orders of magnitude fewer evaluations than the standard MAP-Elites, and that, once solved, the produced encoding can be used for rapid optimization of novel, but similar, tasks. The presented techniques not only scale up quality diversity algorithms to high dimensions, but show that black-box optimization encodings can be automatically learned, rather than hand designed.
During exercise, heart rate has proven to be a good measure in planning workouts. It is not only simple to measure but also well understood and has been used for many years for workout planning. To use heart rate to control physical exercise, a model which predicts future heart rate dependent on a given strain can be utilized. In this paper, we present a mathematical model based on convolution for predicting the heart rate response to strain with four physiologically explainable parameters. This model is based on the general idea of the Fitness-Fatigue model for performance analysis, but is revised here for heart rate analysis. Comparisons show that the Convolution model can compete with other known heart rate models. Furthermore, this new model can be improved by reducing the number of parameters. The remaining parameter seems to be a promising indicator of the actual subject’s fitness.
Analyzing training performance in sport is usually based on standardized test protocols and needs laboratory equipment, e.g., for measuring blood lactate concentration or other physiological body parameters. Avoiding special equipment and standardized test protocols, we show that it is possible to reach a quality of performance simulation comparable to the results of laboratory studies using training models with nothing but training data. For this purpose, we introduce a fitting concept for a performance model that takes the peculiarities of using training data for the task of performance diagnostics into account. With a specific way of data preprocessing, accuracy of laboratory studies can be achieved for about 50% of the tested subjects, while lower correlation of the other 50% can be explained.
The Fitness-Fatigue model (Calvert et al. 1976) is widely used for performance analysis. This antagonistic model is based on a fitness-term, a fatigue-term, and an initial basic level of performance. Instead of generic parameter values, individualizing the model needs a fitting of parameters. With fitted parameters, the model adapts to account for individual responses to strain. Even though in most cases fitting of recorded training data shows useful results, without modification the model cannot be simply used for prediction.