{
  "slug": "prediction-cadence",
  "term": "Prediction cadence",
  "aliases": [
    "scoring cadence"
  ],
  "definition": "Prediction cadence is the schedule or interval at which a prediction system produces successive predictions for the same monitored entity.",
  "note": "Cadence is distinct from telemetry sampling, the historical input window and the future prediction horizon. D-FAD received latency summaries every 20 minutes but scored a rolling 24-hour input window every four hours. Its qualifying alarm-to-replacement window was four hours to 18 days. These are separate clocks, not a recommended monitoring schedule. [goldszmidt-hotstorage12]\n\nThe original evaluation counted unresolved month-end alarms as false positives. Incomplete follow-up does not establish a full event-free horizon; the synthetic counting example explicitly assumes complete observation instead. Repeated scores are also not distinct-drive notifications. [goldszmidt-hotstorage12] [hesela-alert-units]",
  "limitations": [
    {
      "text": "Repeated scores are not automatically separate incident notifications or disk replacements; record the aggregation policy and the unit used by each metric.",
      "citationIds": [
        "hesela-alert-units"
      ]
    }
  ],
  "related": [
    "prediction-horizon",
    "false-positive-rate",
    "precision",
    "recall"
  ],
  "citations": [
    {
      "id": "goldszmidt-hotstorage12",
      "title": "Moises Goldszmidt. Finding soon-to-fail disks in a haystack. HotStorage 2012, sections 3 and 5.",
      "url": "https://www.usenix.org/system/files/conference/hotstorage12/hotstorage12-final48.pdf",
      "accessed": "2026-10-16"
    },
    {
      "id": "hesela-alert-units",
      "title": "Hesela: repeated decisions versus distinct-drive alert counts; synthetic counting example.",
      "url": "https://hesela.com/analyses/ml-disk-failure-prediction/#repeated-predictions",
      "accessed": "2026-10-16"
    }
  ],
  "status": "shipping",
  "advantages": [],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "updated": "2026-10-16"
}