Predicting Campus Placements with Supervised Learning & Skill Gap Scoring

Bridging Academic Curriculums and Industry Standards

Engineering students often struggle to understand which technical competencies (DSA, full-stack, cloud, system design) correlate most strongly with campus recruitment success. In SmartPlacement, I built an analytics engine for university placement cells.

Predictive Modeling & Feature Importance

We analyzed multi-year placement historical records across $500+$ candidates:

1. Primary Predictive Features:

- Semester CGPA & High School academic consistency.

- Verified coding platform ratings (LeetCode, CodeChef, HackerRank).

- Core project complexity score (measured by full-stack repo metrics and live deployments).

- Internship experience (months).

1. Random Forest Feature Importance:

- Project complexity and DSA problem-solving speed outweighed pure CGPA by a factor of $2.4\times$ in tier-1/tier-2 company selection rounds.

// Skill gap scoring calculation engine
interface CandidateMetrics {
  cgpa: number;
  dsaProblemsSolved: number;
  projectsBuilt: number;
  internshipMonths: number;
}

export function computeReadinessScore(metrics: CandidateMetrics): number {
  const dsaWeight = 0.35;
  const projectWeight = 0.30;
  const academicWeight = 0.20;
  const expWeight = 0.15;

  const score =
    Math.min(metrics.dsaProblemsSolved / 250, 1.0) * dsaWeight * 100 +
    Math.min(metrics.projectsBuilt / 5, 1.0) * projectWeight * 100 +
    (metrics.cgpa / 10.0) * academicWeight * 100 +
    Math.min(metrics.internshipMonths / 6, 1.0) * expWeight * 100;

  return Math.round(score);
}

Personalized Roadmap Generator

Outputs dynamic step-by-step roadmaps for students based on their predicted percentile tier.