Skip to content

GRA · Insurance Nexus

Spatial financeWhere capital meets hazard, exposure and transition

Machine learning in insurance

Artificial intelligence and machine learning are used in insurance to read images and documents, detect fraud, triage claims, enrich exposure data and segment risk. On this platform they are treated both as methods and as a source of risk to be governed, tested and explained.

01تۆمارکردن

Machine learning in insurance on the record.

Method recordSTEM method
  1. ڕێگازیرەکی دەستکرد و فێربوونی ماشێن
  2. Applied inInsurance Nexus
  3. داتا4 inputs
  4. Canonical pageSTEM Commons
02بەکارهێنان

What artificial intelligence and machine learning does here.

Artificial intelligence and machine learning develop algorithms that learn patterns from data to classify, predict or generate outputs. They are increasingly used for hazard detection, forecasting and decision support, and are themselves a source of risk. In insurance and catastrophe, climate and hazards and health and biosecurity, machine learning can scale detection and forecasting across interconnected سیستەمەکان. Its own failure modes must be weighed when automated outputs inform consequential decisions. The canonical description is on the STEM Commons.

Exposure enrichment from imagery

Models extract roof type, building footprint and vegetation proximity from aerial and satellite imagery to fill gaps in exposure data, with accuracy measured against inspected properties.

Fraud and claims triage

Classifiers score claims for complexity or fraud indicators to route them for review, with false positive rates, appeal outcomes and group disparities monitored beside detection.

Fairness testing of pricing models

Pricing and underwriting models are tested for disparate outcomes and proxy effects across groups, and less discriminatory alternatives are compared for predictive performance.

03Loss model

Loss exceedance, average annual loss and probable maximum loss.

The loss a portfolio in the footprint can expect per year and at the 1 in 100 and 1 in 250 levels that insurers and lenders price, from the same three blocks every catastrophe model uses. Teaching model with stated illustrative values, not a prediction.

کۆنترۆڵەکان

Typical annual hazard intensity (location)
1.00 intensity index
بڵاوبوونەوەی لێکەوتە توندەکان (پێوەر)
0.80 intensity index
توندییەک کە تێیدا نیوەی بەها لەدەست دەچێت
4.00 intensity index
بڵاوبوونەوەی لاوازی
0.55
بەشی بەها لە شوێنپێی مەترسیدەردا
40%
ستانداردی پاراستن (ماوەی گەڕانەوە)
50 ساڵ

خوێندنەوەکان

تێکڕای زیانی ساڵانە
0.51% of value
زیان لە 1 لە 100 ساڵدا
24.5% of value
زیان لە 1 لە 250 ساڵدا
28.4% of value
توندی لە 1 لە 100 ساڵدا
4.68 intensity index
ژمارە، ڕێگا و سەرچاوەکان
ماوەی گەڕانەوە (ساڵ)Intensity (intensity index)زیان (% ی بەها)
10.090.00
21.270.00
41.910.00
72.470.00
133.000.00
243.510.00
444.020.00
824.5323.55
1545.0326.44
2875.5328.87
5366.0330.88
1,0006.5332.53

ڕێگاکان

  1. مەترسیدەری بەهای زۆرداری Gumbel Gumbel (1958)، Statistics of Extremes، Columbia University Press
  2. ناسکیی لۆگنۆرماڵ (لاوازی) FEMA (2003)، HAZUS-MH Technical Manual
  3. چەمکی ئەگەری تێپەڕبوون و تێکڕای زیانی ساڵانە Grossi و Kunreuther (2005)، Catastrophe Modeling، Springer
  4. کاریگەری بەربەست Di Baldassarre و ئەوانی دیکە (2013)، Socio-hydrology، HESS 17
  5. پەنجەی نادڵنیایی (مۆنتی کارلۆی پارامیتەر) Spiegelhalter, Pearson و Short (2011)، Visualising uncertainty about the future، Science 333
  6. هەستیاریی یەک-یەک (تۆرناد) Saltelli و ئەوانی دیکە (2008)، Global Sensitivity Analysis: The Primer، Wiley

داتا

The model runs on the teaching values set by the controls. To calibrate it for a place, read it against these admitted layers and series, which this page’s map can draw:

  • ڕووبارەکان Natural Earth
  • بەرینی شارەکان Natural Earth
  • شارەکان Natural Earth
  • بوومەلەرزەکانی M4.5+، هەفتەی ڕابردوو (USGS) U.S. Geological Survey · register usgs.earthquake.summary-feeds
  • ڕێژەی باران، نیو کاتژمێری ڕابردوو (IMERG) NASA GPM IMERG Early · register nasa.gibs.context-live-layers
  • پێشبینی ڕۆیشتنی ڕووبار، 30 ڕۆژ لەگەڵ مەودا (GloFAS لە ڕێگەی Open-Meteo) Open-Meteo.com (CC BY 4.0), Copernicus GloFAS · register open-meteo.public-pages
  • بەرینی دروستکراو و نیشتەجێبوون، 2010 (HBASE، Landsat) CIESIN/Columbia, NASA SEDAC HBASE · register sedac.context-urban-layers
Loss exceedance, average annual loss and probable maximum loss. Illustrative teaching model: it runs on the values set by the controls, not on measured data. Methods: Gumbel extreme-value hazard (Gumbel (1958), Statistics of Extremes, Columbia University Press); Lognormal fragility (vulnerability) (FEMA (2003), HAZUS-MH Technical Manual); Exceedance-probability curve and average annual loss (Grossi and Kunreuther (2005), Catastrophe Modeling, Springer).

لە ناو شەش کۆڵەکەکەدا، بە سەرپەرشتیی Nexus Consortium

04داتا

What the method consumes.

Datasets and records used by artificial intelligence and machine learning on this platform, with the role each plays and the label its values carry. Sources are listed in the source register.

داتاڕۆڵناونیشان
Aerial and satellite imageryExposure attributesچاودێریکراو
Labelled claims historiesTraining and testingREPORTED
Property inspection recordsGround truth for imagery modelsچاودێریکراو
Demographic reference dataFairness testingمەزەندەکراو
05Exposure, loss and capital

What machine learning in insurance means for exposure, loss and capital.

The exposure, loss, scenario and capital views a risk officer, supervisor or ministry of finance would ask for, with the frameworks they report against. Structures follow supervisory and actuarial practice. Values are illustrative: never prices, ratings or advice.

Exposure and loss at return periods

How losses from machine learning in insurance grow with severity, gross and net of reinsurance, against capital.

1 in 101 in 501 in 1001 in 250grossnet of reinsurance800
Return periodAnnual probabilityGross lossNet lossNet, capital = 100
1 in 100.10141010
1 in 500.02332121
1 in 1000.01462727
1 in 2500.004683939
تێکڕای زیانی ساڵانە3.12.22.2
Illustrative loss index for a example portfolio (capital = 100): the exceedance-probability form insurers and supervisors read. Not data on any institution, not a price or a rating.
NGFS climate scenarios · Network for Greening the Financial SystemIFRS S2 Climate-related Disclosures · ISSBTCFD recommendations · Financial Stability Board TCFDPrinciples for the effective management of climate-related financial risks · Basel Committee on Banking Supervision
06پشتڕاستکردنەوە

How results are checked.

Models are evaluated on held-out data that respects spatial, temporal and group structure. Robustness, calibration and fairness are tested alongside accuracy.

بپشکنەHow it is done
Temporal and held-out validationModels are tested on later periods and unseen regions, not only on random splits
Calibration and drift monitoringPredicted probabilities are compared with outcomes over time, and drift triggers recalibration
Fairness and explainability reviewOutcome disparities and explanations are reviewed before deployment and periodically after
Human oversightDecisions adverse to customers have a documented human review and an appeal route

Performance degrades when deployment conditions differ from training data. Models can encode biases and fail without clear warning on rare events.

08Stewardship

Who maintains the method.

زیرەکی دەستکرد و سیستەمە خۆسەرەکان stewards artificial intelligence and machine learning for the Consortium and reviews how it is applied across platforms.

09پەیوەست

ئەو شوێنانەی ئەم لاپەڕەیە لە ناو کۆنسۆرتیۆم بۆیان دەبات.

10پرسیارەکان

پرسیار و وەڵام.

What are the limits of artificial intelligence and machine learning in this field?
Performance degrades when deployment conditions differ from training data. Models can encode biases and fail without clear warning on rare events.
How are results checked?
Models are evaluated on held-out data that respects spatial, temporal and group structure. Robustness, calibration and fairness are tested alongside accuracy.
Are the values on this page real?
The data names, roles and checks are real descriptions of practice. The canvas shows a design of the method’s use, not a run; no result is reported here.
Where else is this method applied?
ئەو canonical method page lists every platform that applies it.
Reportsبەشداربە

لەم لاپەڕەیەوە بەڵگە پێشکەش بکە.

Reports is the Consortium’s quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.

Global Risks AllianceInsurance Nexusئامادەیی دارایی و بیمەمەترسی و خۆڕاگریی کەرتی جیهانی
Insurance Nexus

Bring an exposure or capital question to the Alliance.

زانیاری و پرسیار ئەمڕۆ کراوەن. پرسیارێک ژمارەی ئاماژەی پێدەدرێت و هەنگاوی دواتر دیاری دەکات؛ داواکاری، پارەدان، تەرخانکردنی کۆمپیوتەر و تۆمارکردن تەنیا کاتێک دەکرێنەوە کە تیم، مەرج و توانای پێداچوونەوەیان هەبێت.