Valeriy M., PhD, MBA, CQF (@predict_addict)

Update: as some folks have commented it appears to be “calibrated” with a version of Platt’s scaler aka temperature scaling. If true it is a rather dumb idea considering that: 1) Platt’s scaler was invented specifically for support vector machines assuming specific shape of class scores. 2) rel...

Kolmogorov Arnold Attention is all one needs 🧠🔍 KArAt: Learnable (and more explainable) Attention This paper swaps the fixed softmax for Kolmogorov-Arnold Attention (KArAt)—a learnable operator. Why care: • Clearer attention maps: heads often lock onto whole objects, aiding audits. • Interpretabi...

🚀 **The Paper We’ve Been Waiting For!** 🚀 ...

🚀 **Are deep learning models truly necessary for time series forecasting?** 🚀 The paper **"Do We Really Need Deep Learning Models for Time Series Forecasting?"** challenges the common assumption that deep learning—especially Transformer-based models—is the best approach for forecasting....