A collection of heterogeneous distance functions handling missing values.

Interpolation for Missing Data

This project explores the Framingham Heart disease dataset with the objective to predict its risk in 10 years. Various methods for handling missing values and outliers are explored as iterations. After analysing the dataset, important and necessary features are selected. Seven ML models are implemented, with evaluation on the basis of Test Recall.

ML Topics include KNN. Naive Bayes and Support vectors both in Theory and Python Code. KNN Imputation technique is also explained in this branch.

Customer Segmentation with Credit Insights helps businesses personalize marketing and boost retention by analyzing credit card usage patterns. Using K-Means, DBSCAN, and PCA, it identifies meaningful customer segments from high-dimensional transaction data.

Cleaning & preprocessing pipeline for The Met Open Access dataset: consistency/integrity checks + feature engineering + LOF-based outlier detection + KNN imputation for missing values 🌅

A dual-part finance and retail analytics project covering credit default prediction for companies using machine learning (Logistic Regression & Random Forest) and market risk analysis of a five-stock Indian equity portfolio using historical price and return data.

Various data processing techniques for handling missing values

Different imputation technique with example

A Python/Jupyter notebook collection implementing PCA (from scratch and with scikit-learn) and comparing missing-value imputation strategies on toy biological datasets.