Exploring the efficient frontier for massive genome data sets

该项目将开发优化技术，利用大量基因组数据集探索遗传改良和遗传多样性之间的平衡。这些技术的有效实现在很大程度上依赖于数值线性代数，将不得不利用数据的固有属性，并且对于计划的数据集规模，将需要使用高性能计算。对于非常大的数据集，通过精确的方法解决可能是棘手的，在这种情况下，将考虑来自机器学习的启发式技术。

This project will develop optimization techniques to explore the balance between genetic improvement and genetic diversity using massive genome data sets. Efficient implementation of these techniques is largely dependent on numerical linear algebra, will have to exploit inherent properties of the data and, for the planned scale of data sets, will require the use of high-performance computing. For the very largest data sets, solution via exact methods may be intractable, in which case heuristic techniques derived from Machine Learning will be considered.