Karan Singh
Karan Singh
Work History

Contributor

Goolge Summer Of Code · Remote

May 2025 – Present·1 yr 3 mos

Machine Learning Intern

Suvidha Mahila Mandal · Nagpur, Maharashtra

Sep 2024 – Oct 2024·1 mo

In my role as a Machine Learning Intern at Suvidha Foundation, I focused on developing and evaluating NLP models to enhance data processing and summarization capabilities. This involved working with advanced transformer models (BigBird, Longformers, PEGASUS, and more) trained on datasets like CNN/Daily Mail, comparing performance metrics like ROUGE scores to optimize text summarization. It was rewarding to contribute to Suvidha’s mission of educational empowerment, leveraging machine learning to support impactful projects.

 Builds
Wanderer

Wanderer

Sep 14th, 2025

1

Content
GitHub - KaranSinghDev/Fraudulent-Transaction-Detector-And-Analyzer-Using-Gradient-Boosting-Machine: This project implements a fraud detection system using LightGBM, a gradient boosting algorithm. It predicts fraudulent transactions through data preprocessing, feature engineering, and model evaluation. LightGBM efficiently handles imbalanced data, and multiple metrics are used to assess model performance.

GitHub - KaranSinghDev/Fraudulent-Transaction-Detector-And-Analyzer-Using-Gradient-Boosting-Machine: This project implements a fraud detection system using LightGBM, a gradient boosting algorithm. It predicts fraudulent transactions through data preprocessing, feature engineering, and model evaluation. LightGBM efficiently handles imbalanced data, and multiple metrics are used to assess model performance.

This project implements a fraud detection system using LightGBM, a gradient boosting algorithm. It predicts fraudulent transactions through data prepr...

machine learning
GitHub - KaranSinghDev/Fraudulent-Transaction-Detector-And-Analyzer-Using-Gradient-Boosting-Machine: This project implements a fraud detection system using LightGBM, a gradient boosting algorithm. It predicts fraudulent transactions through data preprocessing, feature engineering, and model evaluation. LightGBM efficiently handles imbalanced data, and multiple metrics are used to assess model performance.
 Attended Events
1events attended
DATEEVENTCOMMUNITY
12 Oct 2025
DevFest New Delhi 2025

GDG New Delhi

Communities

GDG New Delhi

59327 members

GDG Cloud New Delhi

52789 members

GDG Noida

38450 members

GDG Gurugram

16158 members

Delhi NCR Kotlin User Group

13606 members

PixelPeps

8275 members

AT A GLANCE

1Builds & Labs
1Events Attended