Tung Anh Tran

Tung AnhTran

Data scientist

Machine learning models for classification and computer vision, on large and complex datasets.

See the projectsGet in touch

Latest write-upReading Tracker: two buttons and a dashboard for my reading habit

In short

I'm Tung Anh Tran, a data scientist building machine learning models for classification and computer vision on large, complex datasets. Most of my work is in healthcare and education, and I'm open to problems in other industries.

More about how I work

Selected work

Competition, DrivenData, 2026

Two views of a brain scan: a calibrated 3D CNN and tree stack for DaT-SPECT

A stacked 3D CNN ensemble and gradient-boosted tree model that separates abnormal from normal DaT-SPECT scans with out-of-fold AUROC 0.9455. The CNN ensemble's probabilities are calibrated to an ECE of 0.0187, and the baseline's calibration carried over to held-out acquisition families at a log-loss cost of +0.0006.

  • medical imaging
  • computer vision
  • binary classification
  • calibration

Read the write-up

0.9455

OOF AUROC, final model

5-fold stratified, n = 1,362

Calibration error (ECE)
0.0187
Calibration penalty, unseen acquisition family
+0.0006
Final private rank
#159 / 1009

  • Hobby project, Reading app, 2026

    Reading Tracker: two buttons and a dashboard for my reading habit

    A hobby app for my own reading. Two buttons log each reading session, and a public dashboard turns the sessions into streaks, pace and an estimate of when I will finish my current book.

  • Project Placeholder

    Project title

    One line that leads with the model and its validated result.

All projects

Published

From the blog

  • Choosing a calibrator out of sample

    Why a calibrator scored on its own training predictions looks perfect.

    Coming soon, 6 min read Placeholder

  • Reading a leaderboard shift

    What a steady gap between public and private scores does and does not say.

    Coming soon, 5 min read Placeholder

  • Post title

    One line on what the reader learns.

    Coming soon, 4 min read Placeholder

All posts

Have a dataset that needs a careful model?

I'm open to data science, machine learning and analytics roles. Tell me about the data and the decision it should support.

Get in touch