Learning with Dependent Data: With Applications in Computational Genome Analysis

经典的机器学习理论建立在独立同分布随机变量的假设之上。然而，在实际应用中，这一假设经常被违反，例如，当数据表现出时间或空间相关性时。在拟议的计划中，我们将为依赖数据学习的理论、高效和有效的学习机器的开发以及其在计算基因组分析中的应用做出贡献。

The classical machine learning theory is built upon the assumption of independent and identically distributed random variables. In practical applications, however, this assumption is often violated, for instance, when the data exhibits temporal or spatial correlations. In the proposed program, we will contribute to the theory of learning with dependent data, the development of efficient and effective learning machines, and the application thereof to computational genome analysis.