Smruti Parmar

About

I am a computer science and engineering student currently pursuing my B.Tech at Dhirubhai Ambani Institute of Information and Communication Technology, with an expected graduation in 2026. I have practical experience as a Software Engineer Research Intern at Dhirubhai Ambani University, where I worked on machine learning and deep AR forecasting. My work involves developing hybrid forecasting models using LSTM, GRU, TCN, and DeepAR architectures to improve anomaly detection in time series data. I have implemented probabilistic forecasting frameworks with Mixture Density Networks and Quantile Regression, and optimized models using specialized loss functions like Tube Loss. I use Python, PyTorch, SQL, and various machine learning techniques to build and optimize predictive systems, including projects on glucose forecasting and calorie burn prediction.

Experience

Software Engineer Research Intern - Machine learning and Deep AR forecasting

Dhirubhai Ambani University

May 2025 – Jul 2025

• Engineered high-performance hybrid forecasting models (LSTM, GRU, TCN, and DeepAR) to improve anomaly detection accuracy in complex, real-world time series datasets. • Implemented Probabilistic Forecasting frameworks using Mixture Density Networks (MDN) and Quantile Regression to quantify uncertainty and identify subtle data irregularities. • Optimized model robustness by integrating specialized loss functions, including Tube Loss, resulting in more precise detection of long-term trend shifts and point anomalies. Executed large-scale experiments and hyperparameter tuning to transform theoretical machine learning concepts into reliable, high-accuracy software solutions.

LSTMDeepARMixture Density Networks

Education

Dhirubhai Ambani Institute of Information and Communication Technology

B.Tech · Information and Communication Technology

2022 – 2026

Skills