School of Engineering & Sciences Faculty

Dr. Apeksha Mittal

Assistant Professor

Qualification: B.Tech (CSE) M.Tech (CSE) Ph.D (CSE)

College/University: Guru Gobind Singh Indraprastha University, Delhi 2013 Banasthali University, Rajasthan 2015 Guru Gobind Singh Indraprastha University, Delhi 2021

Introduction:

Apeksha Mittal has specialized in the area of Artificial Neural Networks. Her Key interest areas are Artificial Neural Networks, Deep Learning, Machine Learning, Artificial Intelligence, Fuzzy Sets, Soft Computing, Algorithms Design and Analysis.

She has research experience in the area of Artificial Neural Networks of 6 years. She has published research papers in journals of international repute (indexing in SCI/ESCI) and various national and international conferences.

Publications (from 2015)

Scholarly Journals

‘Weight and bias initialization routines for Sigmoidal Feedforward Network’, Applied Intelligence, pp 2651-2671.(Mittal, M., Singh, A. P., and Chandra, P., 2021).

‘Improving Learning in Neural Networks through Weight Initializations’, Journal of Information & Optimization Sciences.(Mittal, M., Singh, A. P., and Chandra, P., 2021).

‘A New Weight Initialization using Statistically Resilient Method and Moore-Penrose Inverse based Method for SFANN ’, International Journal of Recent Research Aspects, 4(2), pp 98-105. (Mittal, M., Singh, A. P., and Chandra, P., 2017).

‘A Modification to Nyugen-Widrow Weight Initialization Method’, Intelligent Systems, Technologies and Applications, Advances in Intelligent Systems and Computing, Springer 910, pp. 141-153. (Mittal, M., Singh, A. P., and Chandra, P., 2019).

‘Comparison of Random Weight Initialization to New Weight Initialization CONEXP’, International Conference on Recent Developments in Science, Engineering and Technology, Data Science and Analytics, Springer, 1230, pp. 279-289. (Mittal, M., Singh, A. P., and Chandra, P., 2017).

‘A statistically resilient method of weight initialization for SFANN’, International Conference on Advances in Computing, Communications and Informatics (ICACCI), IEEE, pp. 1371-1376. (Mittal, M., Singh, A. P., and Chandra, P., 2015).

‘Comparison of Statistically Resilient Weight Initialization Method and Random Weight Initialization Method for Benchmark Datasets’, Poster presentation in 3rd Indian Workshop on Machine Learning, IIT(BHU). (Mittal, M., Singh, A. P., and Chandra, P., 2018).

Research

Apeksha Mittal’s research involves development of new weight initialization algorithms to improve the training of Sigmoidal Feedforward Artificial Neural Networks, that has a wide range of applications in the areas of function approximation, predictions, regression and classification.