Data mining technology is an effective way to solve the problem of rich data and poor knowledge. Cluster analysis is an important content of data mining, including analysis ideas based on partition, hierarchy, density, grid and model. Neural network clustering is a typical clustering method based on model thinking. It is an organic combination of brain cognitive science and data mining. It has a strong theoretical connection with actual brain processing knowledge. This paper uses the K-means algorithm to optimize the neural network clustering data mining algorithm, and designs experiments to verify the neural network data mining clustering optimization algorithm proposed in this paper. The experimental research results in this paper show that on four UCI datasets, the mean MP values of the neural network data mining clustering optimization algorithms are 77.9 and 87.72, respectively, which are greater than the values of the other two algorithms. This paper also applies the algorithm to the study of the distribution of remaining oil. The algorithm has achieved obvious results in the cluster analysis of the degree of flooding.

数据挖掘技术是解决数据丰富、知识贫乏问题的有效途径。聚类分析是数据挖掘的重要内容，包括基于分区、层次、密度、网格和模型的分析思路。神经网络聚类是一种典型的基于模型思维的聚类方法。它是脑认知科学与数据挖掘的有机结合。它与实际的大脑处理知识具有很强的理论联系。本文采用K-means算法对神经网络聚类数据挖掘算法进行优化，并设计实验验证本文提出的神经网络数据挖掘聚类优化算法。本文的实验研究结果表明，在四个 UCI 数据集上，神经网络数据挖掘聚类优化算法的平均 MP 值分别为 77.9 和 87.72，均大于其他两种算法的值。本文还将该算法应用于剩余油分布的研究。该算法在洪水程度的聚类分析中取得了明显的效果。