Who I am

Data scientist

I love Data Science, and I'm eager to apply machine/deep learning to solve business problems.

Kaggle expert

I'm actively participating in Kaggle competitions and learning new skills. I currently have one solo silver and two team silvers.

Bioinformatician

I'm exposed to many bioinformatics software and tools. I have a broad experience in NGS analysis and data visualization.

Plant biologist

I have five years experience managing corn populations in both field and greenhouse. I also have extensive molecular biology skills.

Featured work

FireCaster: help animal rescue team save animal lives

Deep learning model to forecast bushfire damage risk on Flinders Chase National Park using satellite imagery and weather data
Three week project at Insight Data Science
[Time-series, Computer Vision, Keras, Satellite Imagery, BigQuery]

Plotly Circos

An open-source software that helps visualize NGS sequencing data with ease.
Live demo is down for maintenance.
Beginner-friendly Bioinformatics Data Visualization Tool
[Data Visualization, Plotly, Dash, React, Docker]

Maize Unstable factor for orange1

My PhD research on an epigentic modifier in maize. I'm a co-first author on a published paper and my contribution includes molecular cloning, generation of transgenic plants, real-time PCR and RNA-seq analysis.
PhD Research
[Thesis Research, Plant breeding, Molecular Cloning, RNA-seq]

Tutorial on RNA-seq data analysis

Reproducible RNA-seq differential expression analysis pipeline
Beginner-friendly RNA-seq tutorial
[RNA-seq, R Markdown, PCA, DESeq, DESeq2, edgeR, limma]

US traffic accident map

Traffic Accident in the US from March 2015 - March 2019
Shiny interactive map
[Data Visualization, R Shiny, shinydashboard, Leaflet]

Kaggle Competition: Toxicity Classification

Build classfication models to predict toxic comments while minimizing bias.
Top 1%, Team Silver
[Kaggle, NLP, PyTorch, Bert, GPT-2, Bi-LSTM]

Kaggle Competition: Elo Merchant Category Recommendation

Build models on unbalanced data to predict customer loyalty score.
Top 3%, Solo Silver
[Kaggle, Feature Engineering, Hyperparameter Tuning, lightgbm]

Kaggle Competition: APTOS 2019 Blindness Detection

Build models on eye images to predict diabetic retinopathy.
Top 4%, Team Silver
[Kaggle, Image Classfication, EfficientNet, Image Augmentation]