Hi, I'm Kesav Kaza
Hi, I'm Kesav Kaza
Hello!! I am a researcher in the field of Electrical Engineering and Computer Science. I am currently working a faculty member at the Indian Institute of Information Technology, Allahabad. I hold a PhD in Electrical Engineering from the Indian Institute of Technology Bombay. After PhD, I worked as a postdoctoral fellow at University of Ottawa and University of Montreal, Canada.
Decision Making under Uncertainty
Human-Automation Teaming
Control Systems
UAV mission and motion planning
Communication Systems and Networks
Stochastic Optimization
Human in the loop systems
Reinforcement Learning
Experiment Design
https://scholar.google.com/citations?user=jBJd7CIAAAAJ&hl=en&oi=ao
Counter Drone Surveillance using Cellular Networks
The commercial availability of low-cost drones in the civilian space has dramatically increased the use of drones across domains including surveillance, delivery, film making, etc. Detection and classification of drones is crucial for security and regulatory purposes, as well as for improving the efficiency of drone operations.
In this project the focus is on detection, tracking of drones using cellular network infras- tructure based on micro-Doppler signatures followed by intent inference based on trajectory, kinematics and local context. A major component of this work includes intent based kinematic modelling. Fusion of data from multiple sensors including human feedback is planned for improving the overall operational efficiency.
Related articles
V. Mehta, K. Kaza, F. Dadboud, M. Bolic and I. Mantegh, “Enhancing Counter Drone Operations Through Human-AI Collaboration: A Hierarchical Decision-Making Framework” presented at the 42nd Digital Avionic Systems Conference, 2023, Barcelona, Spain.
In collaborative human-automation systems it is important to consider the impact on performance, of various human factors such as cognitive workload, fatigue, trust or belief in automation capability, etc.
We consider a scenario where a human-automation team is making potentially risky classification decisions such as labeling a mobile target Hostile or Non-Hostile under time-constraint. The automation observes a batch of independent tasks, analyzes them, and has the option to refer a subset of them to a human expert. However, human performance depends on wor load. We provide algorithms for ranking the tasks according to their risk/uncertainty and find the optimal subset to be referred to the human.
Related articles
K. Kaza, J. Le Ny and A. Mahajan, “Decision Referrals in Human-Automation Teams”, 60th IEEE Conference on Decision and Control (CDC), 2021.
Sequential decision problems are defined in settings where a decision maker interacts with its environment through a set of actions. Based on the action, the environment generates a reward and changes its state. The goal of the decision maker is to maximize its long-term cumulative reward. Restless multi-armed bandits are a class of sequential decision problems in which the environment is composed of seemingly independent (weakly coupled) entities evolving in a Markovian manner (where the next state only depends on the present).
We study two classes of such problems where the environment states are only partially ob- servable by the decision maker (1) lazy restless bandits - where decision instants are sparse compared to the state evolution, (2) constrained restless bandits - where there are availability constraints that are dynamic.
Applications for these models are found in resource allocation in wireless networks, cyber- physical systems, social networks, etc.
Related articles
K. Kaza, R. Meshram, V. Mehta and S. N. Merchant, “Constrained Restless Bandits for Dynamic Scheduling in Cyber-Physical Systems”, 2021, under revision, preprint available at arxiv.org/pdf/1904.08962.pdf
K. Kaza, R. Meshram, V. Mehta and S. N. Merchant, “Sequential Decision Making with Limited Observation Capability: Application to Wireless networks”, in IEEE Transactions on Cognitive Communications and Networking, 2019, DOI: 10.1109/TCCN.2019.2898000.
R. Meshram, and K Kaza. “Simulation Based Algorithms for Markov Decision Processes and multi-action restless bandits”, 2020, under revision, preprint available at arxiv.org/pdf/2007.12933.pdf
R. Meshram, K. Kaza, V. Mehta and S. N. Merchant, “Restless Bandits for Scheduling in Energy Constrained Networks”, 8th IEEE Indian Control Conference (ICC), 2022, Chennai, India.
R. Meshram and K. Kaza, “Indexability and Rollout Policy for Multi-State Partially Observable Restless Bandits”, 60th IEEE Conference on Decision and Control (CDC), 2021.
R. Meshram and K. Kaza, “Monte Carlo Rollout Policy for Recommendation Systems with Dy- namic User Behavior”, IEEE International Conference on Communication Systems and Networks (COMSNETS), 2021.
K.Kaza,V.Mehta,R.MeshramandS.N.Merchant,“RestlessBanditswithCumulativeFeedback: Applications in Wireless networks”, IEEE Wireless Communication and Networking Conference (WCNC), Barcelona, Spain, 2018.
K. Kaza, R. Meshram and S.N.Merchant, “Relay Employment Problem for Unacknowledged Transmissions: Myopic Policy and Structure”, IEEE International Conference on Communica- tions (ICC), Paris, France, 2017.
V. Mehta, K. Kaza, R.Meshram and S.N.Merchant, “Restless Bandits with Constrained Arms: Ap- plication to 5G Small Cell Selection”, IEEE Wireless Communication and Networking Conference (WCNC) - Poster session, Barcelona, Spain, 2018.
V. Mehta, R. Meshram, K. Kaza, S.N. Merchant and U.B. Desai, “Rested and Restless Bandits with Constrained Arms and Hidden States: Applications in Social Networks and 5G Networks”, IEEE Access, vol. 6, pp. 56782-56799, 2018.
V.Mehta, Z. Shaikh, K. Kaza, H. D. Mustafa and S. N. Merchant, “A Crowd-Cloud architecture for Big Data analytics”, Twenty Second National Conference on Communication (NCC), Guwahati, India, 2016.
K. Kaza, K. Kshirsagar and K.S. Rajan, “A Bi-objective Algorithm for Dynamic Reconfiguration of Mobile networks”, In proceedings of IEEE International Conference on Communications (ICC), Ottawa, Canada, 2012.
K.Kshirsagar,K.KazaandK.S.Rajan,“Designof2-levelHierarchicalRingNetworks”,ThirdInter- national Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT), Budapest, Hungary, 2011.
R. Meshram, K. Kaza, “Online Repeated Posted Price Auctions with a Demand Side Platform”, 2019, preprint available at arxiv.org/pdf/1910.01860.pdf