PinnedWalter Sperat·Jul 11, 2023Using optuna with sklearn the right way — Part 1Using optuna with sklearn correctly can get ugly fast, let's try and make it simplerA response icon3A response icon3
Walter Sperat·Apr 5Introducing ForestFire, a new tree-learning libraryWhat is it and why does it exist?
Walter Sperat·Feb 4Pattern Matching for the discerning Data ScientistWhy not just use if/else?A response icon1A response icon1
Walter Sperat·May 27, 2025Our problems just got worse: LLMs can choose to lie to us and there isn't much we can do about itNew research throws light on the problem of aligning LLM's objectives to ours
Walter Sperat·Mar 19, 2025A deep dive into LightGBM — Part 2The train APIA response icon1A response icon1
Walter Sperat·Sep 3, 2024Playing Roulette with Monte CarloUsing simulations to get insights from this game of chance!A response icon2A response icon2
Walter Sperat·Mar 16, 2024Imputing, part 1: Theory behind MissingnessThere's more to missing data than np.nan…
Walter Sperat·Feb 19, 2024The Perilous Path of Laying off Data ScientistsNavigating the Risks of Science