Organic Semiconductor Candidate Molecules Generated via Machine Learning Framework

A 25.7 KB XLSX file containing data from a machine learning framework for designing organic semiconductors. The dataset, authored by Yamato Nakanishi and last updated in June 2026, was generated using hierarchical variational autoencoders, Gaussian mixture regression, and Bayesian optimization to identify molecules with low reorganization energy and high carrier mobility.

Use Cases

• Training generative models for molecular design based on latent space representations described in the paper.
• Benchmarking Bayesian optimization methods for materials discovery using the provided structural descriptors.
• Analyzing the relationship between reorganization energy and carrier mobility in organic semiconductors.
• Generating new candidate molecules by sampling from the described Gaussian mixture model in latent space.

Strengths

• Dataset is focused on a specific, novel application of machine learning for molecular design.
• The generation methodology is described in detail, involving hierarchical VAEs and Bayesian optimization.
• File size is 25.7 KB, indicating a focused and likely manageable dataset.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare, author Yamato Nakanishi.

Collection Method

Generated via a computational framework combining hierarchical variational autoencoders, Gaussian mixture regression, and Bayesian optimization.

Freshness

Last updated 2026-06-04 05:03:39; freshness should be verified.