Hiiii...

github

Umasankar Gudivada  • Peerlist

A simple Product Recommendation System built using Python, scikit-learn, and Streamlit. It suggests similar products based on the user’s search query using a content-based filtering approach.

A Product Recommendation System

This project focuses on predicting customer churn in the telecom industry using machine learning. The dataset contains multi-year subscriber records across regions, service providers, and network technologies. Multiple ML models were trained and evaluated, using different Machine learning models. The model helps identify critical churn patterns such as high-risk circles, providers, and technologies

ChurnMatrix : Telecom Customer Risk Predictio

A Machine learning project to predict credit card delinquency risk using LightGBM, XGBoost, and Logistic Regression. Includes data preprocessing, SMOTE for class imbalance, and model evaluation to help identify high-risk customers for better credit risk management.

PayShield : Credit Delinquency Prediction Mod

A machine learning model predicts diabetes risk using medical data like glucose levels, BMI, and age. It identifies high-risk individuals early to support timely intervention and better healthcare outcomes through accurate and efficient predictions.

RiskRadar : Diabetes

The Speech Translator Web App is an intuitive browser-based tool that captures your voice, converts it to text, and instantly translates it into your desired language — all in real-time. Designed with accessibility and simplicity in mind, the app uses modern web APIs to enable seamless speech recognition and multilingual translation through an interactive and beautifully animated interface. It also supports continuous recognition, animated UI transitions, and language auto-detection, providing a rich and user-friendly experience.

Speech Recognition with Translation

A machine learning project predicting monthly crime rates using FBI data. Includes data preprocessing, lag feature engineering, and modeling with different Machine Learning models, achieving strong accuracy. Great example of time-series forecasting for crime analytics.