Vasilis Kalyvas·Oct 12We always wanted explainable ML models, yet why are we so comfortable with LLMs?A few years ago, whenever I built an ML model, the most normal follow-up question was “why did the model predict this?” And here I was…A response icon1A response icon1
InAI MindbyVasilis Kalyvas·Oct 4“MLshorts” 56: Your model is not very sure. Is it because of aleatoric or epistemic uncertainty?Two similar concepts, but with significant difference.A response icon1A response icon1
InAI MindbyVasilis Kalyvas·Aug 1After ~7 years in Data Science, I’ve realized that model accuracy is rarely the reason a project…Always make the right questions, instead
InAI MindbyVasilis Kalyvas·Jun 8“MLshorts” 55: Imbalanced Classification: What does it really mean to prioritize Precision or…The threshold is the “key”
InAI MindbyVasilis Kalyvas·May 22“MLshorts” 54: Are you overwhelmed by LightGBM parameters?Let’s see how not to be anymore
InAI MindbyVasilis Kalyvas·May 19“MLshorts” 53: Your ML model might be lying to youCross-Validation Explains Why
InAI MindbyVasilis Kalyvas·May 12“MLshorts” 52: Overfitting ⚠️ in Machine LearningHow to spot it before it’s too late
InAI MindbyVasilis Kalyvas·May 10“MLshorts” 51: RMSE Isn’t the Only Metric in Regressionit always has to do with the angle from which you look at your model