Azmain Yakin Srizon | Searcharxiv

arXiv2023

Prognostic Biomarker Identification for Pancreatic Cancer by Analyzing Multiple mRNA Microarray and microRNA Expression Datasets

Possessing the five-year durability rate of nearly 5%, currently, the fourth leading cause for cancer-related deaths is pancreatic cancer. Previously, several works have resolved that early diagnosis performs a meaningful function in enhancing the durability rate and diverse online tools have been utilized to distinguish prognostic biomarker which is a lengthy process. We believe that the statistical feature selection method can produce a better and faster result here. To authenticate our statement, we picked three different mRNA microarray (GSE15471, GSE28735, and GSE16515) and a microRNA (GSE41372) dataset for identification of differentially expressed genes (DEGs) and differentially expressed microRNAs (DEMs). By adopting some feature selecting methods, 178 DEGs and 16 DEMs were elected. After identifying target genes of DEMs, we selected two DEGs (ECT2 and NRP2) which were also identified among DEMs target genes. Moreover, overall durability report established that ECT2 and NRP2 were associated with poor overall survival. Hence, we concluded that for pancreatic cancer, statistical feature selection approaches certainly perform better for biomarker identification than pre-defined online programs, and here, ECT2 and NRP2 can act as possible prognostic biomarkers. All the resources, programs and snippets of our literature can be discovered at https://github.com/Srizon143005/PancreaticCancerBiomarkers.

q-bio.GN↗

arXiv2023

TaLU: A Hybrid Activation Function Combining Tanh and Rectified Linear Unit to Enhance Neural Networks

The application of the deep learning model in classification plays an important role in the accurate detection of the target objects. However, the accuracy is affected by the activation function in the hidden and output layer. In this paper, an activation function called TaLU, which is a combination of Tanh and Rectified Linear Units (ReLU), is used to improve the prediction. ReLU activation function is used by many deep learning researchers for its computational efficiency, ease of implementation, intuitive nature, etc. However, it suffers from a dying gradient problem. For instance, when the input is negative, its output is always zero because its gradient is zero. A number of researchers used different approaches to solve this issue. Some of the most notable are LeakyReLU, Softplus, Softsign, ELU, ThresholdedReLU, etc. This research developed TaLU, a modified activation function combining Tanh and ReLU, which mitigates the dying gradient problem of ReLU. The deep learning model with the proposed activation function was tested on MNIST and CIFAR-10, and it outperforms ReLU and some other studied activation functions in terms of accuracy(upto 6% in most cases, when used with Batch Normalization and a reasonable learning rate).

cs.CV↗