Data Mining based Dimensionality Reduction Techniques
Ayush Soni, Akhtar Rasool, Aditya Dubey, Nilay Khare · 2022 International Conference for Advancement in Technology (ICONAT) · 2022
With the rapid advancement in the science domain, the explosion of available data is seen in recent years. Laboratory instruments are becoming more advanced day by day and are able to capture thousands of measurements in a single experiment therefore dealing with such a huge amount of high dimensional data creates several challenges and leads to data heterogeneity, curse of dimensionality, and inaccuracy. However, a significant portion of this high dimensional data is redundant and can be efficiently reduced to lower dimensions with the help of traditional dimensionality reduction techniques like Principal Component Analysis (PCA) and Linear discriminant Analysis (LDA). Over the last few years, many dimensionality reduction methods are introduced and used as per their application, and it has been observed that some techniques work better for a particular type of real application dataset while do not show much accuracy/efficiency with any other dataset. This paper aims to survey recent approaches which are compared as per their efficiency and accuracy along with their advantages and disadvantages.