Open to opportunities

Shweta Vyas

@shwetavyas1

Message

Data Scientist focused on ML, time-series forecasting, and 99.73% ECG QRS detection accuracy.

India

Message

What I'm looking for

I’m looking for a Data Scientist role where I can build and validate machine learning models with strong feature engineering and time-series forecasting. I want to deliver reliable predictions end-to-end, with opportunities to work on data-driven challenges.

I’m a Data Scientist with a Diploma in Data Science from IIT Madras and a background in Electrical & Electronics Engineering. My work includes automatic QRS complex detection in 12-lead ECG signals, using K-Means clustering with an Average Combined Entropy Criterion, where I achieved 99.73% detection accuracy.

I build practical machine learning pipelines with strong attention to data quality. In a live Kaggle competition, I placed in the top 10% by building a daily cinema attendance forecasting model, engineering lag features and rolling averages, and resolving a data leakage issue before finalizing the ensemble.

From supervised learning and ensemble methods (LightGBM, XGBoost, Random Forest) to time-series forecasting and EDA, I enjoy turning messy data into reliable predictions. I’m also skilled with feature engineering, SQL, and Python’s analytics stack (Pandas, NumPy, scikit-learn, Statsmodels) to deliver results that hold up in validation and real-world benchmarks.

Experience

Work history, roles, and key accomplishments

Cinema Audience Forecasting

IIT Madras

Jan 2025 - Present (1 year 6 months)

Built a forecasting model to predict daily cinema attendance across multiple theatres with varied demand patterns by location, day of week, and season. Achieved R² performance and placed top 10% (261/2,632) on the live Kaggle leaderboard.

Feature Engineering LightGBM XGBoost

Sales Inventory & Retention

IIT Madras

Jan 2025 - Present (1 year 6 months)

Collected, cleaned, and analyzed retail data on sales, inventory, and customer behaviour across product categories. Built seasonal time-series forecasting models, flagged slow-moving stock and anomalies, and produced inventory adjustment recommendations.

Python Time Series Models EDA Data Cleaning Anomaly Detection

QRS Complex Detection

Jai Narain Vyas University

Jan 2018 - Present (8 years 6 months)

Developed and validated an algorithm for automatic QRS complex detection in 12-lead ECG signals using K-Means clustering with an Average Combined Entropy criterion. Achieved 99.73% detection accuracy (17,808/17,856 QRS complexes) on 1,500 single-lead ECGs across 125 patient records.

Matlab Signal Processing ECG QRS Detection Average Combined Entropy Criterion

Education

Degrees, certifications, and relevant coursework

Indian Institute of Technology Madras

Diploma in Data Science, Data Science

2024 - 2025

Completed a Diploma in Data Science at IIT Madras (2024–2025) as part of the BS Degree Program.

Jai Narain Vyas University

Master of Engineering, Power Systems

2014 - 2018

Completed an M.E. in Power Systems (2014–2018), conducting research on automatic QRS complex detection in 12-lead ECG signals using K-Means clustering; achieved 99.73% detection accuracy.