Skip to content
View shivasankari27's full-sized avatar

Block or report shivasankari27

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
shivasankari27/README.md

Hey,I’m Shiva Sankari V A

MSc Integrated Data Science | ML & Healthcare Analytics | Aspiring Data Scientist

I turn messy healthcare data into clear insights, interactive dashboards, and production-ready ML models. I’m building a portfolio that shows I can do the whole pipeline — from raw data → model → deployment — aiming for top-tier internships at companies like Jane Street, Meta, Google, and Stripe.

Quick Hits

  • Current: 4th semester, Integrated M.Sc. Data Science
  • Focus: Machine Learning · Healthcare Analytics · Predictive Modeling · ETL · Deployment
  • Pro: clean code, reproducible notebooks, deployable demos (Streamlit / FastAPI)
  • Goal: land top internships & build impactful healthcare ML products

Tech Stack Languages:Python · SQL
Libraries / ML:NumPy · Pandas · Scikit-learn · Statsmodels
Viz / Dashboards: Matplotlib · Seaborn · Plotly · Streamlit
Deployment / Tools: FastAPI · Docker · Git · GitHub · Jupyter
Other basic PySpark / ETL concepts, time-series tools


Featured Projects Disease Risk Predictor— full ML pipeline predicting disease probability from biomarker data (EDA → feature engineering → model → Streamlit demo).
Patient Readmission Classifier— reduce readmission risk using interpretable models + SHAP explanations.
Medical Cost Forecast- time-series forecasting of treatment costs (ARIMA + LSTM hybrid).
Health Analytics Dashboard — interactive Streamlit dashboard for hospital KPIs and alerts.
Clinical Data Toolkit — reusable cleaning & validation utilities for messy clinical datasets.


✅ What I Care About

  • Reproducibility: seed values, env files, requirements
  • Clarity: short notebooks + modular scripts
  • Impact: metrics that matter (AUC, RMSE, precision/recall, cost savings)
  • Explainability: model interpretation (SHAP / LIME)
  • Deployability: one-click demos with Streamlit / FastAPI + Docker

Currently Learning & Building

  • Advanced feature engineering for clinical time-series
  • Model interpretability (SHAP, counterfactuals)
  • MLOps basics: Dockerized deployments + CI for ML projects
  • Probability & statistics for stronger quant intuition

Connect / Collaborate

Fun Fact I get excited by datasets that look impossible at first glance — the messier, the better. Also, coffee + late-night debugging is my cardio.

Popular repositories Loading

  1. programming-for-everybody programming-for-everybody Public

    Mini project from Coursera Python course

    Python

  2. shivasankari27 shivasankari27 Public

  3. physiobot physiobot Public

    A real-time artificial intelligence system that mimics a therapist by assessing physical therapy actions while off

    Python