COIL2000-train

This is the training set of the COIL 2000 challenge as used by Huang et al. (2020). > Huang, X., Khetan, A., Cvitkovic, M., & Karnin, Z. (2020). > Tabtransformer: Tabular data modeling using contextual embeddings. > arXiv preprint arXiv:2012.06678v1. ## Source: Original Owner and Donor: Peter van der Putten Sentient Machine Research Baarsjesweg 224 1058 AA Amsterdam The Netherlands +31 20 6186927 pvdputten '@' hotmail.com, putten '@' liacs.nl TIC Benchmark Homepage: http://www.liacs.nl/~putten/library/cc2000/ Data Set Information: Information about customers consists of 86 variables and includes product usage data and socio-demographic data derived from zip area codes. The data was supplied by the Dutch data mining company Sentient Machine Research and is based on a real world business problem. The training set contains over 5000 descriptions of customers, including the information of whether or not they have a caravan insurance policy. A test set contains 4000 customers of whom only the organisers know if they have a caravan insurance policy. The data dictionary ([Web Link]) describes the variables used and their values. Note: All the variables starting with M are zipcode variables. They give information on the distribution of that variable, e.g. Rented house, in the zipcode area of the customer. One instance per line with tab delimited fields. TICDATA2000.txt: Dataset to train and validate prediction models and build a description (5822 customer records). Each record consists of 86 attributes, containing sociodemographic data (attribute 1-43) and product ownership (attributes 44-86).The sociodemographic data is derived from zip codes. All customers living in areas with the same zip code have the same sociodemographic attributes. Attribute 86, "CARAVAN:Number of mobile home policies", is the target variable. TICEVAL2000.txt: Dataset for predictions (4000 customer records). It has the same format as TICDATA2000.txt, only the target is missing. Participants are supposed to return the list of predicted targets only. All datasets are in tab delimited format. The meaning of the attributes and attribute values is given below. TICTGTS2000.txt Targets for the evaluation set. Attribute Information: DATA DICTIONARY Nr Name Description Domain 1 MOSTYPE Customer Subtype see L0 2 MAANTHUI Number of houses 1 - 10 3 MGEMOMV Avg size household 1 - 6 4 MGEMLEEF Avg age see L1 5 MOSHOOFD Customer main type see L2 6 MGODRK Roman catholic see L3 7 MGODPR Protestant ... 8 MGODOV Other religion 9 MGODGE No religion 10 MRELGE Married 11 MRELSA Living together 12 MRELOV Other relation 13 MFALLEEN Singles 14 MFGEKIND Household without children 15 MFWEKIND Household with children 16 MOPLHOOG High level education 17 MOPLMIDD Medium level education 18 MOPLLAAG Lower level education 19 MBERHOOG High status 20 MBERZELF Entrepreneur 21 MBERBOER Farmer 22 MBERMIDD Middle management 23 MBERARBG Skilled labourers 24 MBERARBO Unskilled labourers 25 MSKA Social class A 26 MSKB1 Social class B1 27 MSKB2 Social class B2 28 MSKC Social class C 29 MSKD Social class D 30 MHHUUR Rented house 31 MHKOOP Home owners 32 MAUT1 1 car 33 MAUT2 2 cars 34 MAUT0 No car 35 MZFONDS National Health Service 36 MZPART Private health insurance 37 MINKM30 Income < 30.000 38 MINK3045 Income 30-45.000 39 MINK4575 Income 45-75.000 40 MINK7512 Income 75-122.000 41 MINK123M Income >123.000 42 MINKGEM Average income 43 MKOOPKLA Purchasing power class 44 PWAPART Contribution private third party insurance see L4 45 PWABEDR Contribution third party insurance (firms) ... 46 PWALAND Contribution third party insurane (agriculture) 47 PPERSAUT Contribution car policies 48 PBESAUT Contribution delivery van policies 49 PMOTSCO Contribution motorcycle/scooter policies 50 PVRAAUT Contribution lorry policies 51 PAANHANG Contribution trailer policies 52 PTRACTOR Contribution tractor policies 53 PWERKT Contribution agricultural machines policies 54 PBROM Contribution moped policies 55 PLEVEN Contribution life insurances 56 PPERSONG Contribution private accident insurance policies 57 PGEZONG Contribution family accidents insurance policies 58 PWAOREG Contribution disability insurance policies 59 PBRAND Contribution fire policies 60 PZEILPL Contribution surfboard policies 61 PPLEZIER Contribution boat policies 62 PFIETS Contribution bicycle policies 63 PINBOED Contribution property insurance policies 64 PBYSTAND Contribution social security insurance policies 65 AWAPART Number of private third party insurance 1 - 12 66 AWABEDR Number of third party insurance (firms) ... 67 AWALAND Number of third party insurane (agriculture) 68 APERSAUT Number of car policies 69 ABESAUT Number of delivery van policies 70 AMOTSCO Number of motorcycle/scooter policies 71 AVRAAUT Number of lorry policies 72 AAANHANG Number of trailer policies 73 ATRACTOR Number of tractor policies 74 AWERKT Number of agricultural machines policies 75 ABROM Number of moped policies 76 ALEVEN Number of life insurances 77 APERSONG Number of private accident insurance policies 78 AGEZONG Number of family accidents insurance policies 79 AWAOREG Number of disability insurance policies 80 ABRAND Number of fire policies 81 AZEILPL Number of surfboard policies 82 APLEZIER Number of boat policies 83 AFIETS Number of bicycle policies 84 AINBOED Number of property insurance policies 85 ABYSTAND Number of social security insurance policies 86 CARAVAN Number of mobile home policies 0 - 1 L0: Value Label 1 High Income, expensive child 2 Very Important Provincials 3 High status seniors 4 Affluent senior apartments 5 Mixed seniors 6 Career and childcare 7 Dinki's (double income no kids) 8 Middle class families 9 Modern, complete families 10 Stable family 11 Family starters 12 Affluent young families 13 Young all american family 14 Junior cosmopolitan 15 Senior cosmopolitans 16 Students in apartments 17 Fresh masters in the city 18 Single youth 19 Suburban youth 20 Etnically diverse 21 Young urban have-nots 22 Mixed apartment dwellers 23 Young and rising 24 Young, low educated 25 Young seniors in the city 26 Own home elderly 27 Seniors in apartments 28 Residential elderly 29 Porchless seniors: no front yard 30 Religious elderly singles 31 Low income catholics 32 Mixed seniors 33 Lower class large families 34 Large family, employed child 35 Village families 36 Couples with teens 'Married with children' 37 Mixed small town dwellers 38 Traditional families 39 Large religous families 40 Large family farms 41 Mixed rurals L1: 1 20-30 years 2 30-40 years 3 40-50 years 4 50-60 years 5 60-70 years 6 70-80 years L2: 1 Successful hedonists 2 Driven Growers 3 Average Family 4 Career Loners 5 Living well 6 Cruising Seniors 7 Retired and Religeous 8 Family with grown ups 9 Conservative families 10 Farmers L3: 0 0% 1 1 - 10% 2 11 - 23% 3 24 - 36% 4 37 - 49% 5 50 - 62% 6 63 - 75% 7 76 - 88% 8 89 - 99% 9 100% L4: 0 f 0 1 f 1 - 49 2 f 50 - 99 3 f 100 - 199 4 f 200 - 499 5 f 500 - 999 6 f 1000 - 4999 7 f 5000 - 9999 8 f 10.000 - 19.999 9 f 20.000 - ? ## Past Usage P. van der Putten and M. van Someren (eds). [CoIL Challenge 2000: The Insurance Company Case](http://www.liacs.nl/~putten/library/cc2000/report2.html). Published by Sentient Machine Research, Amsterdam. Also a Leiden Institute of Advanced Computer Science Technical Report 2000-09. June 22, 2000. In this report you will find 29 short papers and extended abstracts on this problem. Acknowledgements Data is (c) Sentient Machine Research 2000 This dataset is owned and supplied by the Dutch datamining company Sentient Machine Research, and is based on real world business data. You are allowed to use this dataset and accompanying information for non commercial research and education purposes only. It is explicitly not allowed to use this dataset for commercial education or demonstration purposes. Please cite/acknowledge: P. van der Putten and M. van Someren (eds) . CoIL Challenge 2000: The Insurance Company Case. Published by Sentient Machine Research, Amsterdam. Also a Leiden Institute of Advanced Computer Science Technical Report 2000-09. June 22, 2000. References and Further Information There is a special website for this benchmark at http://www.liacs.nl/~putten/library/cc2000/. On the website you can find an online report featuring 29 papers written by participants in the CoIL Challenge 2000 and further background information. In future more papers will be added to the website. If you have any submissions, please send them to putten@liacs.nl. ## Note * This is only the training set. UCI also provides the test set. * The [coil dataset](https://openml.org/d/298) contains the full dataset but fails to correctly code categorical variables.