Population-Based Black-Box Optimization for Biological Sequence Design
Christof Angermüller, David Belanger, Andreea Gane, Zelda Mariet, David Dohan, Kevin Murphy, Lucy J. Colwell, D. Sculley
Abstract
SUMMARY
The Upper Silesian Coal Basin (USCB) in Poland is one of the most seismically active mining areas in the world. Seismic observations supported by the Central Mining Institute date back to the 1950s, and include an extensive mine tremors data bank, which has been in operation since that time. More than 55 900 mine tremors of energy E≥ 105 J (local magnitude ML≥ 1.5) occurred over the period 1974–2005.
In the USCB, two types of tremors can be distinguished. These are the so-called ‘mining-tectonic’ seismicity and ‘mining’ seismicity. The ‘mining-tectonic’ seismicity results from the interaction between mining and tectonic factors. These seismic events appear to be located in tectonically disturbed zones and their sources are visibly more energetic. The mine tremor source mechanism can mostly be the normal dip-slip fault with a marked strike-slip component in the source region. The rupture plane azimuths and dips of these seismic events correlate well with the strike and dip of the pre-existing local faults, in the vicinity of which the mine tremor sources have been located. The generalized mine tremor source mechanism of the events under consideration contributed a less than 15 per cent implosion component, an about 15 per cent uniaxial compressional component, and an about 70 per cent shear component to the total seismic moment tensor solution.
The second one—‘mining seismicity’—is strongly associated with the mining activity and seismic events that occur in the vicinity of active mining excavations. These events are energetically weak and their source mechanism can be of explosive type. The total seismic moment tensor is composed of: 20–50 per cent explosive component, practically the same amount of uniaxial compressional component and a very small amount of shear component (less than 10 per cent).
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Cited by top-tier papers21
- Accelerating Bayesian Optimization for Biological Sequence Design with Denoising AutoencodersSamuel Stanton, Wesley J. Maddox, Nate Gruver, Phillip M. Maffettone et al.ICML 2022 · 137 citations
- Design-Bench: Benchmarks for Data-Driven Offline Model-Based OptimizationBrandon Trabucco, Xinyang Geng, Aviral Kumar, Sergey LevineICML 2022 · 126 citations
- Conservative Objective Models for Effective Offline Model-Based OptimizationBrandon Trabucco, Aviral Kumar, Xinyang Geng, Sergey LevineICML 2021 · 119 citations
- Proximal Exploration for Model-guided Protein Sequence DesignZhizhou Ren, Jiahan Li, Fan Ding, Yuan Zhou et al.ICML 2022 · 52 citations
- Data-Driven Offline Optimization for Architecting Hardware AcceleratorsAviral Kumar, Amir Yazdanbakhsh, Milad Hashemi, Kevin Swersky et al.ICLR 2022 · 45 citations
Builds on1
Related papers
- Dynamic Multi-Network Mining of Tensor Time SeriesKohei Obata, Koki Kawabata, Yasuko Matsubara, Yasushi SakuraiWWW 2024 · 13 citations
- MiningVis: Visual Analytics of the Bitcoin Mining EconomyNatkamon Tovanich, Nicolas Soulié, Nicolas Heulot, Petra IsenbergIEEE VIS 2021 · 18 citations
- Unsupervised Event Chain Mining from Multiple DocumentsYizhu Jiao, Ming Zhong, Jiaming Shen, Yunyi Zhang et al.WWW 2023 · 11 citations
- Multi-Source Uncertainty Mining for Deep Unsupervised Saliency DetectionYifan Wang, Wenbo Zhang, Lijun Wang, Ting Liu et al.CVPR 2022 · 58 citations
- Unearthing InSights into Mars: Unsupervised Source Separation with Limited DataAli Siahkoohi, Rudy Morel, Maarten V. de Hoop, Erwan Allys et al.ICML 2023 · 6 citations
