Sunia T.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

I help organizations turn complex data into accurate forecasts, predictive models, and actionable insights using machine learning, statistics, and applied mathematics. I currently work as a Research Scientist at the University of Michigan and hold a dual PhD in Mechanical Engineering and Computational Mathematics. My expertise includes time-series forecasting, statistical modeling, machine learning, predictive analytics, signal processing, uncertainty quantification, and decision-support systems. I specialize in challenging datasets where noise, missing information, uncertainty, or complex system behavior make standard approaches unreliable. What I can help you with: • Time-series forecasting and predictive analytics • Machine learning model development and evaluation • Statistical analysis and uncertainty quantification • Predictive modeling for business, research, and engineering decisions • Classification on difficult or noisy datasets • Feature engineering, data preprocessing, and model validation • Model debugging and performance improvement • End-to-end ML pipelines in Python • Signal processing and biomedical data analysis • Simulation, surrogate modeling, and optimization workflows • Decision-support models for real-world systems Experience Highlights: • Developed machine-learning models for biomedical signal classification that improved accuracy by 15% over conventional feature-engineering approaches. • Developed custom statistical and machine-learning algorithms for identifying patterns, data drift, and behavioral changes in noisy, uncertain, and stochastic datasets. • Built predictive models and surrogate modeling systems that replaced computationally expensive simulations, enabling rapid design evaluation and optimization. • Conducted research in forecasting, stochastic systems, machine learning, and scientific computing, resulting in peer-reviewed publications, open-source software contributions, and conference presentations. Tools & Technologies: Python, Pandas, NumPy, SciPy, scikit-learn, XGBoost, TensorFlow, PyTorch, SQL, Jupyter, Matplotlib, Statistical Modeling, Machine Learning, Time-Series Forecasting, Predictive Modeling, Signal Processing, Data Analysis, Feature Engineering, Model Validation, Optimization, Scientific Computing. My Approach: Most machine-learning problems do not fail because of the model itself - they fail because the data is noisy, incomplete, poorly structured, or difficult to interpret. My focus is on building reliable models, validating assumptions, reducing overfitting, and delivering results that are useful for real-world decisions. If you're working on forecasting, predictive modeling, machine learning, statistical analysis, model validation, or complex data challenges, I’d be happy to discuss how I can help.