Damanjeet Singh

Electronics & Communication Engineering graduate (2026) working in applied machine learning for radar signal intelligence. Built an open-set emitter classification pipeline in C++17 at DRDO-DLRL, benchmarked across 15 dataset partitions and 3 openness levels.

AI/ML Intern - Defence Electronics Research Laboratory (DLRL), DRDO - Hyderabad, India

(2026-02)

• Engineered an open-set radar emitter classification pipeline in C++17, pairing a Random Forest with a multi-criterion rejection stage over 4 pulse-descriptor features.
• Benchmarked rejection logic across 15 dataset partitions and 3 openness levels against learned-fusion baselines, replacing single-split comparison with paired significance testing.
• Audited the codebase and diagnosed 4 correctness defects inflating reported accuracy, including label leakage and train-set contamination, then re-validated every result.
• Traced a post-expansion accuracy regression to feature dilution and introduced a discriminability pre-test now applied before every feature change.
• Evaluated HDBSCAN clustering with DTOA-guided merging to deinterleave overlapping pulse streams, verifying reproducible inference across 3 dataset-difficulty tiers.
• Authored a technical report and an IEEE-format manuscript spanning methodology, a 15-partition evaluation protocol, ablations, and stated limitations.

B.E. - Electronics and Communication Engineering - Thapar Institute of Engineering and Technology (2022-08 - 2026-06) Patiala, India