Microarray data analysis is a major challenging field of research in recent days. Machine learning–based automated gene data classification is an essential aspect for diagnosis of gene related any malfunctions and diseases. As the size of the data is very large, it is essential to design a suitable classifier that can process huge amount of data. Deep learning is one of the advanced machine learning techniques to mitigate these types of problems. Due the presence of more number of hidden layers, it can easily handle the big amount of data. We have presented a method of classification to understand the convergence of training deep neural network (DNN). The assumptions are taken as the inputs do not degenerate and the network is over-parameterized. Also the number of hidden neurons is sufficiently large. Authors in this piece of work have used DNN for classifying the gene expressions data. The dataset used in the work contains the bone marrow expressions of 72 leukemia patients. A five-layer DNN classifier is designed for classifying acute lymphocyte (ALL) and acute myelocytic (AML) samples. The network is trained with 80% data and rest 20% data is considered for validation purpose. Proposed DNN classifier is providing a satisfactory result as compared to other classifiers. Two types of leukemia are classified with 98.2% accuracy, 96.59% sensitivity, and 97.9% specificity. The different types of computer-aided analyses of genes can be helpful to genetic and virology researchers as well in future generation.

We propose a hybridized technique named Spatial-Spectral-Superpixelwise PCA-based Dense 2D-3D CNN Fusion Architecture (3SPCA-D-2D-3D-CNN), that deals with the detailed and complex study of dimensionality reduction and classification of Hyperspectal images (HSI). Our work is 2-fold: At first (1), 3SPCA is applied on raw HSI that adopts superpixels-based local reconstruction to filter the images, whereas PCA-based supplementary global features acquire the relevant and low-dimensional local features. Every HSI pixel is reconstituted by the pixels of its closest neighbors in the same superpixel block to reduce noise and improve spatial information. Next, PCA is conducted on every zone and the full HSI to get local and global features. The local-global and spatial-spectral properties are then concatenated. Secondly (2), the D-2D-3D-CNN fusion architecture is made up of three 3D convolution blocks, two 2D convolution blocks with varied kernel sizes and filters, and four fully connected (FC) dense layers, totaling nine distinguishing and information-enriched features. These features can generate precise class labels and apply them to the appropriate landcovers. The proposed method has been applied to three publicly available HSI landcover datasets, the Indian Pines, the Salinas Valley, and the Pavia University. It achieved respectively 98.33%, 99.99%, and 98.73% average accuracy scores. Due to its improved Feature Extraction capacity from a limited number of training samples and its classification performance with fewer epochs, this method outperforms other relevant state-of-the-art CNN-based methods.