News
- Our paper has been accepted to CVPR 2022 (A* Conference):
- Abhishek Kumar, Oladayo S. Ajani, Swagatam Das and Rammohan Mallipeddi.-GridShift: A Faster Mode-Seeking Algorithm for Image Segmentation and Object Tracking
- Our paper has been accepted to IEEE Transactions on Cybernatics (SCI IF-19.118):
- Abhishek Kumar, Swagatam Das, Lingping Kong and Vaclav Snasel.-Self-Adaptive Spherical Search With a Low-Precision Projection Matrix for Real-World Optimization
- Our paper has been accepted to Swarm and Evolutionary Computation (SCI IF-10.267):
- Abhishek Kumar, Partha P. Biswas and Ponnuthurai N. Suganthan.-Differential evolution with orthogonal array‐based initialization and a novel selection strategy
... - We are organizing "Special Session & Competition on Single Objective Bound Constrained Optimization"
- at GECCO-2022, Boston, USA, 9-13 July 2022 (with 8/4 or 2-page paper),
- at WCCI-CEC-2022, Padova, Italy, 18-23 July 2022 (with full paper) .
- We are organizing "Special Session & Competition on Real-World Multiobjective Constrained Optimization"
- at CEC-2021, Krakow, Poland, 28 June -01 July 2021 (with full paper),
- at GECCO 2021, Lille, France, 10-14 July 2021 (with 2-page paper).
- We are organizing "Special Session & Competition on Single Objective Bound Constrained Optimization"
- at CEC-2021, Krakow, Poland, 28 June -01 July 2021 (with full paper),
- at GECCO 2021, Lille, France, 10-14 July 2021 (with 2-page paper).
- Our paper has been accepted to IEEE Transactions on Smart Grid (SCI IF-10.275):
- Abhishek Kumar, Swagatam Das, and Rammohan Mallipeddi.-An Inversion-free Robust Power Flow Algorithm for Microgrids.
- Our paper has been accepted to IEEE Transactions on Cybernatics (SCI IF-19.118):
- Abhishek Kumar, Swagatam Das, Rakesh Kumar Misra, and Deveneder Singh.-A v-constrained Matrix Adaptation Evolution Strategy with Broyden-based Mutation for Constrained Optimization.
- Our paper has been accepted to IEEE Transactions on Cybernatics (SCI IF-19.118):
- Abhishek Kumar, Swagatam Das, and Rammohan Mallipeddi.-A Reference Vec-tor based Simplified Covariance Matrix Adaptation Evolution Strategy for Con-strained Global Optimization.
Short Biography
Dr. Abhishek Kumar received the B.Tech degree in Electrical Engineering from Uttarakhand Technical University, Dehradun in 2013. He finished his Ph.D. in Systems Engineering at the Department of Electrical Engineering, Indian Institute of Technology (BHU), Varanasi, India in 2019. He is the recipient of the “Young Researcher Award-2016” from the IEEE CIS Chapter, UP section, IIT Kanpur. His developed optimization algorithms “EBOwithCMAR” and “SASS” have secured the first position in IEEE CEC-2017 special session and competition on bound-constrained optimization and IEEE CEC-2020/GECCO-2020 special session and competition on real-world constrained optimization, respectively. His current research interests include swarm and evolutionary computation and its application in real-world optimization problems especially in Power System Optimization applications and Machine Learning. He also serves as a reviewer for several journals including IEEE TCYB, IET GTD, SWEVO, and ASOC.
Research Interest
My research is mainly centered around designing simple yet effective methods to address a variety of Machine Learning Problems, Theoretical Analysis of Learning Algorithms, Optimization Theory, and Application of Machine Learning and Optimization Algorithms in the field of Computer Vision and Modern Power System.
- Unsupervised and Semi-supervised Learning: Our major research directions involve the development of enhanced variants of traditional clustering algorithms to combat the influence of noise in the datasets. As well as that, we also propose applications of these clustering algorithms to unsupervised and semi-supervised training of computer vision problems including image segmentation and object detection that can also be addressed in the context of deep learning and provide effective solutions.
- Theoretical Machine Learning: As part of our research in this domain, we aim to demonstrate the convergence of clustering and classification algorithms, to investigate how probability divergence measures can be used in machine learning algorithms, etc.
- Evolutionary Optimization: Secondly, we are looking into the design and development of innovative variants of evolutionary optimization techniques. These variants may be effective for a range of problems such as those associated with high detrimental, noisy, large scale, and so forth. As well as the application of such optimizer to problem solving, we are also interested in exploring their theoretical analysis and exploring their applicability to various machine learning problems. For example, improving the performance of a classifier, analyzing adversarial attacks on deep classifiers, modern power system problems etc.
I am also interested in the following research areas.
- Constraint Handling Techniques,
- Multi- and Many-objective Optimization,
- Large-scale Optimization,
- Application of Optimization Algorithms in Modern Power System problems.