Big Data for Insurance Companies

Marine Corlosquet-Habart is a certified actuary and co-director of EURIA (Euro-Institut d'Actuariat, University of West Brittany, Brest, France).Jacques Janssen is Honorary Professor at the Solvay Business School (ULB) in Brussels, Belgium.

• Foreword xiJean-Charles POMEROLIntroduction  xiiiMarine CORLOSQUET-HABART and Jacques JANSSENChapter 1. Introduction to Big Data and Its Applications in Insurance 1Romain BILLOT, Cécile BOTHOREL and Philippe LENCA1.1. The explosion of data: a typical day in the 2010s 11.2. How is big data defined?  41.3. Characterizing big data with the five Vs  51.3.1. Variety  61.3.2. Volume  71.3.3. Velocity   91.3.4. Towards the five Vs: veracity and value 91.3.5. Other possible Vs 111.4. Architecture 111.4.1. An increasingly complex technical ecosystem 121.4.2. Migration towards a data-oriented strategy 171.4.3. Is migration towards a big data architecture necessary? 181.5. Challenges and opportunities for the world of insurance  201.6. Conclusion  221.7. Bibliography   23Chapter 2. From Conventional Data Analysis Methods to Big Data Analytics 27Gilbert SAPORTA2.1. From data analysis to data mining: exploring and predicting  272.2. Obsolete approaches 282.3. Understanding or predicting? 302.4. Validation of predictive models 302.4.1. Elements of learning theory 312.4.2. Cross-validation  342.5. Combination of models 342.6. The high dimension case 362.6.1. Regularized regressions 362.6.2. Sparse methods  382.7. The end of science? 392.8. Bibliography   40Chapter 3. Statistical Learning Methods  43Franck VERMET3.1. Introduction  433.1.1. Supervised learning 443.1.2. Unsupervised learning  463.2. Decision trees  463.3. Neural networks  493.3.1. From real to formal neuron 503.3.2. Simple Perceptron as linear separator  523.3.3. Multilayer Perceptron as a function approximation tool 543.3.4. The gradient backpropagation algorithm 563.4. Support vector machines (SVM)  623.4.1. Linear separator  623.4.2. Nonlinear separator 663.5. Model aggregation methods 663.5.1. Bagging  673.5.2. Random forests  693.5.3. Boosting   703.5.4. Stacking   743.6. Kohonen unsupervised classification algorithm  743.6.1. Notations and definition of the model  763.6.2. Kohonen algorithm 773.6.3. Applications  793.7. Bibliography   79Chapter 4. Current Vision and Market Prospective 83Florence PICARD4.1. The insurance market: structured, regulated and long-term perspective 834.1.1. A highly regulated and controlled profession 844.1.2. A wide range of long-term activities  854.1.3. A market related to economic activity  874.1.4. Products that are contracts: a business based on the law  874.1.5. An economic model based on data and actuarial expertise  884.2. Big data context: new uses, new behaviors and new economic models 894.2.1. Impact of big data on insurance companies  904.2.2. Big data and digital: a profound societal change 914.2.3. Client confidence in algorithms and technology 934.2.4. Some sort of negligence as regards the possible consequences of digital traces   944.2.5. New economic models  954.3. Opportunities: new methods, new offers, new insurable risks, new management tools  954.3.1. New data processing methods  964.3.2. Personalized marketing and refined prices 984.3.3. New offers based on new criteria  1004.3.4. New risks to be insured 1014.3.5. New methods to better serve and manage clients 1024.4. Risks weakening of the business: competition from new actors, “uberization”, contraction of market volume 1034.4.1. The risk of demutualization 1034.4.2. The risk of “uberization” 1044.4.3. The risk of an omniscient “Google” in the dominant position due to data 1054.4.4. The risk of competition with new companies created for a digital world 1054.4.5. The risk of reduction in the scope of property insurance  1064.4.6. The risk of non-access to data or prohibition of use  1074.4.7. The risk of cyber attacks and the risk of non-compliance  1084.4.8. Risks of internal rigidities and training efforts to implement  1094.5. Ethical and trust issues 1094.5.1. Ethical charter and labeling: proof of loyalty 1104.5.2. Price, ethics and trust 1124.6. Mobilization of insurers in view of big data  1134.6.1. A first-phase “new converts”  1134.6.2. A phase of appropriation and experimentation in different fields  1154.6.3. Changes in organization and management and major training efforts to be carried out  1184.6.4. A new form of insurance: “connected” insurance 1184.6.5. Insurtech and collaborative economy press for innovation  1214.7. Strategy avenues for the future 1224.7.1. Paradoxes and anticipation difficulties  1224.7.2. Several possible choices 1234.7.3. Unavoidable developments 1274.8. Bibliography   128Chapter 5. Using Big Data in Insurance  131Emmanuel BERTHELÉ5.1. Insurance, an industry particularly suited to the development of big data 1315.1.1. An industry that has developed through the use of data  1315.1.2. Link between data and insurable assets  1365.1.3. Multiplication of data sources of potential interest  1385.2. Examples of application in different insurance activities  1415.2.1. Use for pricing purposes and product offer orientation   1425.2.2. Automobile insurance and telematics  1435.2.3. Index-based insurance of weather-sensitive events 1455.2.4. Orientation of savings in life insurance in a context of low interest rates 1465.2.5. Fight against fraud 1485.2.6. Asset management 1505.2.7. Reinsurance  1505.3. New professions and evolution of induced organizations for insurance companies  1515.3.1. New professions related to data management, processing and valuation 1515.3.2. Development of partnerships between insurers and third-party companies 1535.4. Development constraints 1535.4.1. Constraints specific to the insurance industry 1535.4.2. Constraints non-specific to the insurance industry  1555.4.3. Constraints, according to the purposes, with regard to the types of algorithms used  1585.4.4. Scarcity of profiles and main differences with actuaries  1595.5. Bibliography   161List of Authors  163Index 165