scikit-learn-pipelines

A collection of pandas & scikit-learn compatible transformers for preprocessing and feature engineering 🛠

Optimizing an ML Pipeline in Azure - A Machine Learning Engineer Project

Data Science case for windmill power prediction based on weather. Based on Data Challenge of Air Liquide (@AirLiquide) and TotalEnergies(@total-sa, @Total-RD) companies in 2021. Real Data Science model was much more complicated to get 6 place. The link of the competition - https://datascience.total.com/fr/challenge/19/details#.

[Template] To Create MCPStack MCP Tools in ~5 command lines!

A reproducible, leak-free machine learning benchmarking lab for regression model comparison, cross-validation, diagnostics, experiment tracking, and interactive Streamlit-based inference using the California Housing dataset.

Programming assignments covering fundamentals of machine learning and deep learning. These were completed as part of the Plaksha Tech Leaders Fellowship program.

End-to-end ML and MLOps pipeline for Diabetes Prediction including model training, FastAPI API serving, Docker containerization, and production-ready deployment workflow

Production-grade B2B SaaS churn prediction system with leak-free Scikit-Learn pipelines, custom ROI/MRR financial scorers, calibrated probabilities, FastAPI serving, and PSI/KS drift monitoring.

Build custom vacab, Ham /Spam using tfidf , Movie review classification using TFIDF

Analysis on effect of temperature, rainfall and pesticide on Global food crops yield along with Crop Yield prediction using scikit-learn ML algorithms. Interactive website for outcomes using HTML, Bootstrap, JS, CSS, d3 and so on.

A production-grade Explainable AI (XAI) framework for Credit Risk Assessment. Uses LightGBM and SHAP to transform "black-box" loan default predictions into transparent, human-readable insights. High-performance FastAPI implementation designed for scalable financial risk modeling and real-time interpretability.

A Modular, Production-Style ML Pipeline with Class-Imbalance Handling

This is Learning process and implementing of Machine Learning Model and learning never stop

Code repository for "MyeVAE: A multi-omics variational autoencoder for predicting mortality in newly diagnosed myeloma patients"

This Repo contains the Mini-Projects I've made while self-learning Data Science.

Apprenticeship project with Indian Institute of Science. Objectives were to Model tumor and immune cell interaction in the presence of MMP & TIMP by varying their diffusion rates and check the effect of MMP8 on these interactions. These simulations provides insights into how ECM remodelling shapes cancer progression.

Benchmark of spatial validation and conformal prediction for water-quality data. On two open groundwater-arsenic datasets (20,575 Punjab wells; 56,175 global pixels) it measures how random cross-validation overstates skill and how interval and exceedance-probability calibration degrade under regional shift.

Unsupervised ML project segmenting mall customers into 6 behavior-based groups using StandardScaler, PCA, and K-Means — with business recommendations derived from each segment.

Stroke risk prediction using Random Forest, XGBoost, SVM & Logistic Regression with SMOTE for class imbalance