Amirtha Ganesh R

Dual-engine synthetic tabular data generator using Gaussian Copula and CTGAN with automated preprocessing and comprehensive evaluation metrics.

ML-powered fitness platform analyzing 20K+ records using PCA, K-Means clustering, and neural networks for personalized insights.

Water quality intelligence system with dual XGBoost models for WQI regression and classification deployed as interactive web app.

Sequential neural networks (RNN, LSTM, GRU) for text classification with real-time prediction on news articles and reviews.

Classifies news articles as Fake or Real using TF-IDF features for classical models and LSTM embeddings for deep learning.

YOLOv8-based livestock detection system for identifying and counting sheep in aerial imagery. mAP50: 0.782, Precision: 0.793.

Comparative study of YOLOv8 Nano at different resolutions (416×416 vs 608×608) for garbage detection. Analyzes speed-accuracy trade-offs.

Data-driven approach to identifying optimal locations for Quick Service Restaurant expansion in Maharashtra using demographic and POI data.

Framework to identify high-potential markets and operational risks in e-commerce expansion. Delivered in 4 hours using 9 interconnected datasets.

Identifying engines operating at abnormally high turbine gas temperatures using Isolation Forest and unsupervised machine learning.

Forecasting electrical load demand and solar power generation in Italy using ARIMA models for grid management and energy planning.

Analyzing 80,000 student records with linear and non-linear models. R²: 0.8705, previous GPA correlation: 0.93.

Deep learning project classifying handwritten digits (0–9) from 28×28 grayscale images. Validation Accuracy: ~97%.

Collection of DL experiments: LSTM text classification, MLP regression, RNN sentiment analysis, CNN image classification on CIFAR-10.

Full-stack financial forecasting pipeline with 116 engineered features. Linear Regression outperforms XGBoost/LightGBM in out-of-sample accuracy.

Predicting hospital readmission using ensemble learning on 101,766 patient records. Gradient Boosting achieves 69.82% accuracy.

Analyzing relationship between trader profitability and Bitcoin Fear & Greed Index using 211K+ trades. Fear periods yield 2.5× higher PnL.