Krish Kalsi
Data science student
About
(01)Third-year data science student at NMIMS, Mumbai. I build models whose claims can be checked — prediction intervals, simulations, forecasts — and I’m most at home where statistics meets markets.
2024–28 · CGPA 8.43
Quant finance · Econometrics
NHANES survey-design audit
Summer 2027
Work
06Research
(03) · Working papers- [P1]2026
Kalsi, K. et al. “Autonomous Stock Trading: A Comparative Review.”
Systematic review · In final revision for submission · Supervisor: Dr Prashasti Kanikar, NMIMS MPSTME
A review of 19 papers from four databases, screened against criteria fixed in advance. It compares supervised price-prediction models with deep reinforcement-learning trading agents, and finds that reward functions rarely account for risk and that almost no study tests whether simulated results carry over to live trading.
- [P2]2026–27
Kalsi, K. et al. “Survey Design in Medical Machine Learning: An Audit of NHANES-Based Studies.”
Supervised research · In progress · Supervisor: Dr Prashasti Kanikar, NMIMS MPSTME
NHANES is a complex survey with strata, clusters and sampling weights, yet much published machine-learning work treats it as a simple random sample. We measure how common this is and what it does to the reported results.
- [P3]2025
Kalsi, K. et al. “Critical Path Method for Quick-Commerce Logistics: Bottleneck Identification in Last-Mile Gig Delivery.”
Working paper
Models a last-mile delivery run as an Activity-on-Node network (CPM/PERT) and proposes an average negative float measure that locates where delays build up.
Experience
