Research

Research Highlights

• Nanotechnology: Synthesized and optimized nanoparticles for bio-detection systems and flexible electronics.
• Data Analytics & Machine Learning: Applied machine learning models to predict nanoparticle properties and optimize material designs for smart textiles and bio-detection.
• Smart Textiles: Developed flexible antennas and energy-harvesting devices for wearable electronics using advanced materials and computational models.

Machine Learning Projects in Research

1. Gold Nanoparticle Size and Surface Prediction

Objective: Develop a predictive model to determine the size and surface properties of gold nanoparticles (Au NPs) based on growth factors. Model: Random Forest and Gradient Boosting

• Problem: Accurately predict Au NP size and surface properties using growth factors such as temperature, time, and precursor concentrations.
• Approach: Trained Random Forest and Gradient Boosting models on experimental data to identify key features affecting nanoparticle size.
• Results: Achieved a high R² score of 0.997 and RMSE of 0.348. Feature importance analysis showed that precursor concentration and temperature were the top contributors.

2. Future Baby Names Prediction: Leveraging Famous Influences with Machine Learning

Objective: Predict future baby name trends based on historical data and cultural influences using a logistic regression model. Model: Logistic Regression

• Problem: Predicting popular baby names using time-series data and demographic information, addressing overfitting and underfitting issues.
• Approach: Developed a logistic regression model with 98% accuracy, employing feature clustering to identify key influencers (celebrity names, historical figures).
• Results: Created a 5-year prediction of baby name trends with insights for marketing and demographic forecasting.

Research Assistant | Biophysics Lab | North Carolina State University

JAN 2017 – AUG 2018, RALEIGH, NC

• Experimental: Conducted research on DNA bio-detection using optical anisotropy of upconversion nanoparticles (UCNPs) to probe intricate protein-DNA interactions. Enhanced the accuracy of bio-detection systems by optimizing nanoparticle formulations.
• Data Analysis: Analyzed time-series data from single-molecule spectroscopy to study protein motion dynamics. Applied statistical models and data visualization techniques, leading to better understanding of molecular interactions.

Research Assistant | Smart Textile Research | North Carolina State University

SEP 2018 – MAR 2022, RALEIGH, NC

• Experimental: Led collaboration with the ASSIST Center to develop flexible textile-based antennas for wearable electronics. Applied advanced spectroscopy techniques (FTIR, Raman) to optimize ink formulations, improving conductivity and durability by 20 wt.%.
• Data-Driven: Leveraged machine learning models (Random Forest, multivariate analysis) to optimize nanomaterial-based inks. Conducted data analysis using Python and MATLAB to assess antenna flexibility and predict performance under various conditions.

Research Assistant | Donghua University

SEP 2013 – MAR 2016, SHANGHAI, CHINA

• Experimental: Synthesized Ag nanoparticles and applied them in bio-detection systems. Developed spray-coating techniques for nanoparticle deposition, enhancing LSPR effects, which led to a 4-fold increase in bio-detection signal strength.
• Data Analysis: Conducted UV-Vis and fluorescence spectroscopy testing, applying regression analysis to evaluate the effects of nanoparticle interparticle distance on signal intensity. Optimized synthesis using ANOVA and regression models.

Additional Contributions:

• Published 11 research papers and presented findings at 10+ conferences in the fields of nanotechnology, wearable electronics, and bio-detection systems.
• Applied machine learning and data analytics to enhance experimental design and predict material properties in research projects across smart textiles and bio-nanotechnology.
• Led interdisciplinary collaborations and contributed to the development of market-ready products in wearable electronics and biosensor technologies.

Skills and Tools:

• Data Analysis & Machine Learning: Python (Pandas, NumPy, Matplotlib), R, MATLAB, Scikit-learn
• Algorithms: Random Forest, Gradient Boosting, Decision Trees, Logistic Regression, K-Means Clustering
• Spectroscopy Techniques: FTIR, Raman, UV-Vis, Fluorescence Spectroscopy
• Statistical Analysis: ANOVA, Time-Series Data, Signal Processing, Regression Analysis