Update README.md

README.md

@@ -1,3 +1,86 @@
----
-license: mit
----
+---
+license: mit
+language:
+- en
+pretty_name: ABX-CT-010 Cycling Strategy Collapse
+tags:
+- antibiotics
+- stewardship
+- cycling
+- resistance
+- mdr
+- tabular
+task_categories:
+- tabular-classification
+size_categories:
+- n<1k
+---
+
+ABX-CT-010 Cycling Strategy Collapse
+
+Purpose
+
+Detect when antibiotic rotation stops controlling resistance.
+
+Core pattern
+
+- stress_index high
+- cycle_exposure_index high
+- burden_gap stays high across different cycle drugs
+- mdr_marker rises and stays high
+- later clinical_failure_flag appears
+
+Files
+
+- data/train.csv
+- data/test.csv
+- scorer.py
+
+Schema
+
+Each row is one cycle timepoint in a within series rotation history.
+
+Required columns
+
+- row_id
+- series_id
+- timepoint_d
+- cycle_id
+- drug_in_cycle
+- stress_index
+- cycle_exposure_index
+- population_resistance_burden
+- expected_burden_under_cycling
+- burden_gap
+- mdr_marker
+- mic_to_active_drug_fold_vs_baseline
+- clinical_failure_flag
+- source_type
+- cycling_collapse_signal
+- earliest_cycling_collapse
+
+Labels
+
+- cycling_collapse_signal
+  - 1 for rows at or after first confirmed collapse point
+
+- earliest_cycling_collapse
+  - 1 only for the first collapse row in that series
+
+Scorer logic in v1
+
+- candidate collapse point
+  - stress_index at least 0.80
+  - cycle_exposure_index at least 0.80
+  - burden_gap at least 0.25 at two consecutive points
+  - consecutive points must be different drug_in_cycle values
+  - mdr_marker at least 0.60 and up at least 0.10 vs baseline
+  - exclude burden_gap spike then snapback artifacts
+- confirmation
+  - clinical_failure_flag equals 1 later in series
+
+Evaluation
+
+Run
+
+- python scorer.py --path data/test.csv