Adaptive Feature Selection and Image Classification Using Manifold Learning Techniques
Amna Ashraf, Nazri Mohd Nawi, Muhammad Aamir
Abstract
Manifold learning technique aims to the non-linear dimension reduction of data. The foundation of this concept is the notion that high dimensionality of data features is required to be reduced. Dimension reduction is the field of interest and demand of many data analysts. Moreover it is widely used in computer vision, image processing, pattern recognition, neural networks and machine learning. The research has been divided in to two phases to recognize manifold learning techniques’ importance. In the first phase, manifold learning approach is used to improve the ‘feature selection by clustering’. Clustering algorithms such as K-means, spectral clustering and Gaussian Mixer Model have been tested with manifold learning approaches for adaptive feature selection and the results obtained are satisfactory as compared to simple clustering. In the second phase, a Triple Layered Convolutional Architecture (TLCA) has been proposed for image classification bearing 85.34%, 59.14%, 71.43%, 90.06% and 71.71% accuracy level for the datasets such as Pistashio, Animal, HAR, Mango Leaves and Cards respectively. The performance of proposed TLCA model is compared to the other deep learning models i.e. CNN, LSTM and GRU. In order to further improve the accuracy, reduced dimensional data from manifold learning technique is used and achieved higher accuracies from Hybrid Triple Layered Convolutional Architecture HTLCA 97.73%, 87.18%, 97.97%, 99.19% and 96.91% for the above mentioned sequence of datasets. The effectiveness and precision of the suggested methods are demonstrated by the experimental findings.