Based-on-Hadoop-K-means-clustering-algorithm-Parallel-implementation

Project Description

Use the Hadoop platform and MapReduce programming model to complete the parallel implementation of the K-means clustering algorithm under big data, and apply it to the tourist travel city preference analysis of cities in China The project mainly consists of three parts

• Big data K-means parallel clustering（hadoop_kmeans folder）
• K-means parallel clustering application（application_hadoop_kmeans folder）
• Tourist preference Page display（web_hadoop_kmeans folder）

Big data K-means parallel clustering

Using hadoop's mapreduce programming model, through the K-means clustering algorithm, complete the cluster analysis of 4 million pieces of data, and realize it in the following three environments:

• Stand-alone
• pseudo-distributed
• fully distributed

K-means parallel clustering application

Use hadoop's mapreduce programming model, through the K-means clustering algorithm, and use the crawled ticket data to analyze the tourist preferences of provincial capital cities

• Airline data preprocessing
• Cluster analysis of all capital cities

Tourist preference Page display

Use the Python Flask framework and Baidu Map API to analyze the tourist city preferences of tourists in various cities

• The more times you go to a city, the bigger the point
• You can select the corresponding city through the drop-down box