The Application And Research Of Data Mining In Data Analysis Of Targeted Poverty Alleviation

Targeted poverty alleviation is a way to help people lift themselves out of poverty.According to conditions of different areas and different poor farmers,It takes scientific and effective procedures to actualize accurately identification,assistance and management.In the current work of poverty alleviation,poor identification and help measures matching are all through the visit of village cadres and filing riser.However,there are more than 600 million rural population in China,traditional methods are time-consuming and hard to manage.It is urgent to introduce new methods,which can reduce labor cost and human disturbance and make poverty alleviation work intelligent and transparent.This paper introduces the data mining technology to the poverty alleviation work.Through a series of data processing and analysis programs,it can distinguish the poverty characteristics of poor households' intelligently,and it also lay the foundation for precisely matching assistance measures.First,data preparation,which includes a data index system and data pre-processing.Through the in-depth study of the problem,policies,programs of poverty alleviation,combining the preliminary analysis of the internal data,such as the filing card,the visit data of village cadres,Internet data,health data,civil data,education data,etc,we design a data index system of poverty alleviation under the consideration of expert opinion.It makes a great difference for data to be stored and managed scientifically and effectively so that serving for further data analysis.Then,in the data mining,we use DBSCAN clustering algorithm to mine the poverty characteristics of the poor households.For the high-dimensional and large amount data for poverty alleviation,this paper proposes several improved strategies ofDBSCAN algorithm to improve its clustering accuracy and efficiency : we first use the binary local sensitive hashing(LSH)which can map the data into low dimensional space and enable faster region query for the k neighbors of a data point.The binary data representation method based on k neighborhood is then proposed to map the dataset into the Hamming space for faster cluster expansion.We define a core point based on binary influence space to enhance the robustness to various densities.Also,we propose a seed point selection method,which is based on influence space and k neighborhood similarity,to select some seed points instead of all neighborhood during cluster expansion.Consequently,the number of region queries can be decreased.The experimental results show that the improved algorithm can greatly improve the clustering speed under the premise of ensuring better algorithm clustering accuracy,especially for large-scale datasets.Finally,we apply the data mining scheme designed in this paper to analyze the poverty alleviation data of one prefecture level city,which include four districts,three counties and one city.Through the data pre-processing and DBSCAN algorithm,we can identify the poverty characteristics of different poor households intelligently.Then,we exhibit the analysis results both macroscopically and microcosmically.It provides further decision support for matching support measures and help cadres in the regional poverty alleviation work.