Towards Generative Design of Computationally Efficient Mathematical Models with Evolutionary Learning

Kalyuzhnaya, Anna V.; Nikitin, Nikolay O.; Hvatov, Alexander; Maslyaev, Mikhail; Yachmenkov, Mikhail; Boukhanovsky, Alexander V.

doi:10.3390/e23010028

Cited by 14 publications

(10 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the cable dome support joint provides rigid support for the entire cable dome structure, topology optimization selects the maximum stiffness (minimum flexibility) as the goal. To achieve maximum stiffness with minimum material, the volume fraction of the joint is taken as the constraint, and the design variable of the model is the element density [ 24 , 25 ]. Therefore, the mathematical model of topology optimization can be described as follows: min C ( x ) = 1/2 U T KU s.t.…”

Section: Topology Optimization Methods and Mathematical Modelmentioning

confidence: 99%

Innovative Joint for Cable Dome Structure Based on Topology Optimization and Additive Manufacturing

Wang

Zhu

et al. 2021

Materials

View full text Add to dashboard Cite

Aiming at the problems of a low material utilization rate and uneven stress distribution of cast-steel support joints in cable dome structures, topology optimization and additive manufacturing methods are used for optimization design and integrated manufacturing. First, the basic principle and calculation process of topology optimization are briefly introduced. Then, the initial model of the support joint is calculated and analyzed by using the universal software ANSYS Workbench 2020R2 and Altair OptiStruct, and the optimized joint is imported into Discovery Live to smooth the surface. The static behaviors of three types of joints (topology-optimized joints, joints after the smoothing treatment, and joints from practical engineering) are compared and analyzed. Finally, the joints are printed by using fused deposition modeling (FDM) technology and laser-based powder bed fusion (LBPBF) technology in additive manufacturing. The results show that the new support joint in the cable dome structure obtained by the topology optimization method has the advantages of a novel shape, a high material utilization rate, and a uniform stress distribution. Additive manufacturing technology can allow the manufacture of complex shape components with high precision and high speed. The combination of topology optimization and additive manufacturing effectively realizes the advanced design and integrated manufacturing of support joints for cable dome structures.

show abstract

Section: Topology Optimization Methods and Mathematical Modelmentioning

confidence: 99%

Innovative Joint for Cable Dome Structure Based on Topology Optimization and Additive Manufacturing

Wang

Zhu

et al. 2021

Materials

View full text Add to dashboard Cite

show abstract

“…Every experiment consists of several stages: the evaluation of each version of the evolutionary algorithm (repeated ten times to obtain a stable result); evaluation of the prediction quality (with test sample) and computational time (measured during the evaluations and verified using an empirical performance model [4]); analysis of Pareto frontiers and hypervolume values. For the non-parameter-free algorithms, the population size was set to 20 and the maximum number of generations was set to 30.…”

Section: A Setup Of Experimentsmentioning

confidence: 99%

“…In many cases, it is impossible to maximize all criteria simultaneously, which leads to a multi-objective optimization problem. The usage of multi-objective approaches in AutoML solutions is a quite promising direction that can lead to better suitability of the obtained ML pipelines and even makes it possible to co-design the models and infrastructure [4].…”

Section: Introductionmentioning

confidence: 99%

Multi-Objective Evolutionary Design of Composite Data-Driven Models

Polonskaia

Nikitin

Revin

et al. 2021

2021 IEEE Congress on Evolutionary Computation (CEC)

Self Cite

View full text Add to dashboard Cite

In this paper, a multi-objective approach for the design of composite data-driven mathematical models is proposed. It allows automating the identification of graph-based heterogeneous pipelines that consist of different blocks: machine learning models, data preprocessing blocks, etc. The implemented approach is based on a parameter-free genetic algorithm (GA) for model design called GPComp@Free. It is developed to be part of automated machine learning solutions and to increase the efficiency of the modeling pipeline automation. A set of experiments was conducted to verify the correctness and efficiency of the proposed approach and substantiate the selected solutions. The experimental results confirm that a multi-objective approach to the model design allows achieving better diversity and quality of obtained models. The implemented approach is available as a part of the open-source AutoML framework FEDOT.

show abstract

“…Fine-tuning of the hyperparameters in the composite pipeline usually differents from that are using in existing Au-toML tools for single machine learning models [45]. As an example, the data preprocessing block also require appropriate tuning, but the quality metric evaluation can not be provided without connection with the models.…”

Section: Issue 5: How To Tune the Hyperparameters In Composite Pipeli...mentioning

confidence: 99%

Automated Evolutionary Approach for the Design of Composite Machine Learning Pipelines

Nikitin,

Vychuzhanin,

Sarafanov

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The effectiveness of the machine learning methods for real-world tasks depends on the proper structure of the modeling pipeline. The proposed approach is aimed to automate the design of composite machine learning pipelines, which is equivalent to computation workflows that consist of models and data operations. The approach combines key ideas of both automated machine learning and workflow management systems. It designs the pipelines with a customizable graph-based structure, analyzes the obtained results, and reproduces them. The evolutionary approach is used for the flexible identification of pipeline structure. The additional algorithms for sensitivity analysis, atomization, and hyperparameter tuning are implemented to improve the effectiveness of the approach. Also, the software implementation on this approach is presented as an open-source framework. The set of experiments is conducted for the different datasets and tasks (classification, regression, time series forecasting). The obtained results confirm the correctness and effectiveness of the proposed approach in the comparison with the state-of-the-art competitors and baseline solutions.

show abstract

Towards Generative Design of Computationally Efficient Mathematical Models with Evolutionary Learning

Cited by 14 publications

References 40 publications

Innovative Joint for Cable Dome Structure Based on Topology Optimization and Additive Manufacturing

Innovative Joint for Cable Dome Structure Based on Topology Optimization and Additive Manufacturing

Multi-Objective Evolutionary Design of Composite Data-Driven Models

Automated Evolutionary Approach for the Design of Composite Machine Learning Pipelines

Contact Info

Product

Resources

About