normative
Time-series splits must respect chronology
Time-series data should be split by time, training through some year and testing forward, because splitting across unrelated sectors or random dates lets the model learn from future correlations, which corrupts the test.
normativepending
Speaker
JeremiahEvidence Quote
“you want to split up by time. So, you want to train through some year, you want to test from that year forward”
Created: 11/11/2026, 7:15:49 PM
Your notes
Loading notes...