cnai-ds,github

Hi👋, Feel free to reach me out for freelance ML/Data Science any collaborative research projects via LinkedIn

I am a Data Scientist with strong curiosity, autonomy, creativity and learning ability. Currently, I specialize in pricing optimization using data science and have experience in remote sensing, uplift modeling, marketing promotion optimization based on CTLV, frequentist/Bayesian statistical tests, and strategic consulting.

Visualization samples

cnai-ds's Projects

crop-yield-prediction-satellite-image

Python data pipeline to acquire, clean, and calculate vegetation indices from Sentinel-2 satellite image. Package is available only for our clients.

datathon-criteo-benchmark-churn-date

Datathon with a retargeting ad company. Churn date prediction (Normalized RSME 38), clustering (98% Silhouette), automated identification of the gaps between best and average client within a cluster.

datathon-health-risk-prediction

SCOR Datathon in 2020. Acquired and processed open data, predicted level of Glycohemoglobin, Cholesterol and probability of diabetes, then identified the probability change with Random Survival Forest to suggest improvements to a user.

datathon-instagram-image-color-analysis

Launchmetrics Datathon 2020. Identified the impact of Instagram Image color property to company revenues with OpenCV using HSV and RGB color space.

experimentation-product-analysis

Demonstration of experimentation analysis approach, including A/B testing, A/A testing, and quasi-experimental testing for data driven product decision making.

extremely-imbalanced

Extremely imbalanced binary classification for Travel Insurance Claim. Seeking for an improvement. Usage of Several models including Sequential Neural Network with Back Propagation, Calibration curve, Brier score

nlp-qa-forum

Prediction of the best answers of the questions using a scraped QA forum data, including the texts. Neural Network (linear stack of layers): Accuracy 92%, MSE 0.08

premium-prediction-natural-hazards

Prediction of market premiums for property damage and business interruption insurance products. Added natural hazard data and stacked 3 best models as the final model.