Manh Tuan's Blog
Practical notes on machine learning, AI, and data analysis, with real code and numbers you can reproduce.
Latest posts
-
Labels beat models: what I learned from the RSNA knee MRI competition
A solo run in a 6,425-team Kaggle competition: what I built, why a 0.958 public score became 0.942 private, and what the top teams did differently.
-
Same F1 score, different winner: picking a credit card fraud model
Logistic regression, a decision tree, and a neural network on 284,807 card transactions with only 0.17% fraud. Two models tied on F1, and catching more fraud decided the winner.