Analysis of Suitable Approaches for Data Mining Algorithms

Data mining is the knowledge discovery method by examining the huge bulks of information from numerous perspectives and summarizing it into valuable data; data mining has become an important component in numerous fields. It is used to recognize hidden patterns in a huge data set. In this paper, we are using three techniques of Data mining i.e. classification, clustering, and regression. The paper helps users to identify the best algorithm suitable for their dataset along with their advantages and disadvantages. It also shows the accuracy of the best four algorithms for classification and regression, and for clustering, it shows the number of clusters for different clustering algorithms. The performance of algorithms depends on the size of datasets. As the size of the dataset increases, the performance will also increase. This will reduce the work of users finding the best algorithm using different tools like weka, orange, TPOT, etc. The advantages and disadvantages can help users identify the single best algorithm for further analysis. This system also helps the user to identify target columns from the dataset and various dispensable columns in the dataset.