Dataset Source: https://machinehack.com/hackathon/predict_accident_risk_score_for_unique_postcode/data

Data set is for private consumption for the competition.

According to IBEF “Domestic automobiles production increased at 2.36% CAGR between FY16-20 with 26.36 million vehicles being manufactured in the country in FY20.Overall, domestic automobiles sales increased at 1.29% CAGR between FY16-FY20 with 21.55 million vehicles being sold in FY20”.The rise in vehicles on the road will also lead to multiple challenges and the road will be more vulnerable to accidents.Increased accident rates also leads to more insurance claims and payouts rise for insurance companies.

In order to pre-emptively plan for the losses, the insurance firms leverage accident data to understand the risk across the geographical units e.g. Postal code/district etc.

In this challenge, we are providing you the dataset to predict the “Accident_Risk_Index” against the postcodes.Accident_Risk_Index (mean casualties at a postcode) = sum(Number_of_casualities)/count(Accident_ID)

Working example:

Train Data (given)
Accident_ID Postcode Number_of_casualities 1 AL1 1JJ 2 2 AL1 1JP 3 3 AL1 3PS 2 4 AL1 3PS 1 5 AL1 3PS 1 Modelling Train Data (Rolled up at Postcode level)
Postcode Derived_feature1 Derived_feature2 Accident_risk_Index AL1 1JJ _ _ 2 AL1 1JP _ _ 3 AL1 3PS _ _ 1.33 The participants are required to predict the 'Accident_risk_index' for the test.csv and against the postcode on the test data.

Then submit your 'my_submission_file.csv' on the submission tab of the hackathon page.

Pro-tip: The participants are required to perform feature engineering to first roll-up the train data at postcode level and create a column as “accident_risk_index” and optimize the model against postcode level.

Few Hypothesis to help you think: "More accidents happen in the later part of the day as those are office hours causing congestion"

"Postal codes with more single carriage roads have more accidents"

(***In the above hypothesis features such as office_hours_flag and #single _carriage roads can be formed)

Additionally, we are providing you with road network data (contains info on the nearest road to a postcode and it's characteristics) and population data (contains info about population at area level). This info are for augmentation of features, but not mandatory to use.

The provided dataset contains the following files:

Train: 4,84,042 rows x 27 columns

Test: 1,15,958 rows x 27 columns

train.csv & test.csv:

'Accident_ID', 'Police_Force', 'Number_of_Vehicles', 'Number_of_Casualties', 'Date', 'Day_of_Week', 'Time', ‘Local_Authority_(District)', 'Local_Authority_(Highway)', '1st_Road_Class', '1st_Road_Number', 'Road_Type', 'Speed_limit', '2nd_Road_Class', '2nd_Road_Number', 'Pedestrian_Crossing-Human_Control', 'Pedestrian_Crossing-Physical_Facilities', 'Light_Conditions', ‘'Weather_Conditions', 'Road_Surface_Conditions', 'Special_Conditions_at_Site', 'Carriageway_Hazards', 'Urban_or_Rural_Area', 'Did_Police_Officer_Attend_Scene_of_Accident', 'state', 'postcode', 'country'

Population: 8,035 rows x 10 columns

population.csv:

​​'postcode', 'Rural Urban', 'Variable: All usual residents; measures: Value', 'Variable: Males; measures: Value', 'Variable: Females; measures: Value', ‘Variable: Lives in a household; measures: Value', ‘Variable: Lives in a communal establishment; measures: Value', 'Variable: Schoolchild or full-time student aged 4 and over at their non term-time address; measures: Value', 'Variable: Area (Hectares); measures: Value', 'Variable: Density (number of persons per hectare); measures: Value'

Road Network: 91,566 rows x 8 columns

roads_network.csv:

'WKT', 'roadClassi', ‘roadFuncti', 'formOfWay', 'length', 'primaryRou', 'distance to the nearest point on rd', 'postcode’

Overview Swiss Re is one of the largest reinsurers in the world headquartered in Zurich with offices in over 25 countries. Swiss Re’s core expertise is in underwriting in life, health, as well as the property and casualty insurance space whereas its tech strategy focuses on developing smarter and innovative solutions for clients’ value chains by leveraging data and technology.

The company’s vision is to make the world more resilient. Swiss Re believes in applying fresh perspectives, knowledge and capital to anticipate and manage risk to create smarter solutions and help the world rebuild, renew and move forward.About 1300 professionals that work in the Swiss Re Global Business Solutions Center (BSC), Bangalore combine experience, expertise and out-of-the-box thinking to bring Swiss Re's core business to life by creating new business opportunities.

This dataset is now static. It is therefore no longer updated and is out of date.

This dataset is a product of a national assessment of flood risk for England produced using local expertise. This dataset is produced using [Risk of Flooding from Rivers and Sea] which shows the chance of flooding from rivers and/or the sea, based on cells of 50m. Each cell is allocated one of four flood risk categories, taking into account flood defences and their condition.

This dataset uses OS address data and Royal Mail postcode data to show how many properties are in each of four flood risk categories in each postcode, based simply on the category allocated to the cell that each property is in.

Energy Performance Certificates (EPCs) are needed whenever a property is built, sold or rented. An EPC contains information about a property's energy use and typical energy costs and recommendations about how to reduce energy use and save money. An EPC gives a property an energy efficiency rating from A (most efficient) to G (least efficient) and it is valid for 10 years. The Standard Assessment Procedure (SAP) used to create the EPC is the methodology used by the Government to assess and compare the energy and environmental performance of dwellings. It aims to provide accurate and reliable assessments of dwelling energy performances that are needed to underpin energy and environmental policy initiatives. The data come from an IBM Fuel Poverty report and provide SAP/EPC energy rating by post code within the Glasgow Housing Association (GHA) stock register. The fields are: Post Code, Current Energy Efficiency Rating, Potential Energy Efficiency Rating, Current Environmental Impact Rating and Potential Environmental Impact Rating. Date extracted 2011-05-19. Data supplied by Glasgow Housing Association Licence: None

Geodemographic classifications are small area classifications of social, economic and demographic characteristics. The Output Area Classification (OAC) is a free geodemographic classification. It is an Office of National Statistics validated measure that summarises neighbourhood conditions at the Output Area Level across the United Kingdom. Linkage of this valuable statistics has been problematic for users more used to address records that are georeferenced using unit postcodes. OACoder resolves this problem by allowing users to link corresponding OAC codes to each of the postcode addresses. OACoder is an open source software, and it is developed and tested to work on different versions of windows operating systems. It is stored in Figshare. The source code of the OACoder is stored in SourceForge. As open source software, OACoder has reuse potential across a range of applications. The functionality of OACoder can be extended to work with new version of OAC (2011 OAC). It is also possible to reuse the source code and extend the functionality to work on different operating systems other than Windows. Different components of the software can be reused for the purpose of reading/writing CSV files and handling large data sets.

This software is made available under a GPL-3.0 license, and is described in the following paper: Muhammad Adnan, Alex Singleton, Paul A. Longley. 2013. OACoder: Postcode Coding Tool. Journal of Open Research Software, 1(1) DOI: http://dx.doi.org/10.5334/511ba2c94d661

Code-Point® with polygons shows the notional shape of every postcode unit in Great Britain, and includes major buildings with multiple postcodes. For compelling visuals, Code-Point with polygons lets you apply shading to individual postcodes on a map. This means you can analyse location data at the most granular level and bring your results vividly to life. We give you every single postcode in Great Britain and Northern Ireland – including those for different floors of high-rise buildings. For accuracy, we give every postcode a positional quality rating and map out the boundaries of only the postcodes we can locate most precisely. Code-Point® with polygons contains postcode boundaries for Great Britain. These show the extent of each postcode unit, enabling you to analyse information by postcode. Ideal for activities such as sales targeting or market profiling, as well as any statistical work. Includes notional polygons; vertical streets data; postcode units; eastings and northings; NHS® health authority codes; administrative codes; PO box indicator; and types of delivery points.

Abstract copyright UK Data Service and data collection copyright owner.

The Acorn geodemographic classification is a long-running classification developed by CACI Limited. Acorn operates by merging geography with demographics and details about consumer characteristics and behaviours. Supported by advanced AI methods, comprehensive input data, and detailed product literature, Acorn provides precise information and enables an in-depth understanding of the different types of consumers in every part of the country.

The current classification groups the entire United Kingdom population into 7 categories, 22 groups and 65 types. The data is available at unit postcode level. Further information may be found on the CACI ACORN microsite.

Use of the data requires approval from the data owner or their nominee and is restricted to those based at a Higher Education or Further Education institution. Please see the Data Access section for further information.

For the second edition (October 2024) data and documentation files for 2024 have been added to the study.

Geoplan mapping data provided the flood risk insurer with most accurate UK Postcode data on the market, giving them the ability to create flood maps, catastrophe models and analytics, used by some of the world's largest insurers.

A best-fit lookup between postcodes, 2021 Census Output Areas (OA), Workplace Zones (WZ), Lower Layer Super Output Areas (LSOA), Middle Layer Super Output Areas (MSOA) and current local authority districts (LAD) along with OA, WZ, and LAD classifications as at August 2024 in the UK. Postcodes are best-fitted by plotting the location of the postcode's mean address into the areas of the output geographies. (File size 44MB).Field Names - PCD7, PCD8, PCDS, DOINTR, DOTERM, USERTYPE, OSEAST1M, OSNRTH1M, OA21CD, OAC11CD, OAC11NM, WZ11CD, WZC11CD, WZC11NM, LSOA21CD, LSOA21NM, MSOA21CD, MSOA21NM, LADCD, LADNM, LADNMW, LACCD, LACNMField Types - All TextField Lengths - 7, 8, 8, 6, 6, 1, 6, 7, 9, 3, 48, 9, 2, 60, 9, 63, 9, 35, 9, 36, 35, 3, 48

Open Flood Risk by Postcode is derived from the Environment Agency's Risk of Flooding from Rivers and Sea which allocates a risk level to areas in England, UK. Using postcode data from Open Postcode Geo, each English postcode is placed in its risk area, allowing a flood risk level to be allocated to a postcode.

Fields

• postcode
• FID
• PROB_4BAND
• SUITABILITY
• PUB_DATE
• RISK_FOR_INSURANCE_SOP
• easting
• northing
• latitude
• longitude

PROB_4BAND is the flood risk level, and can be one of the folowing:

• High
• Medium
• Low
• Very Low
• None

Note that where a postcode is outside a flood risk area, some of the column values will be NULL, represented as \N in this file.

Documentation

You can find full documentation on the Open Flood Risk by Postcode homepage.

Acknowlegements

Derived from Risk of Flooding from Rivers and Sea Derived from Open Postcode Geo Licensed under the OGL

Foto von Luke Moss auf Unsplash

This method returns Census 2021 estimates that classify households by occupancy rating based on the number of rooms in the household.

Occupancy rating for rooms defines whether a household's accommodation is overcrowded, ideally occupied or under-occupied. This is calculated by comparing the number of rooms the household requires to the number of available rooms.

The number of rooms the household requires uses a formula which states that:

• one-person households require three rooms comprised of two common rooms and one bedroom
• two-or-more person households require a minimum of two common rooms and a bedroom for each person inline with the Bedroom Standard

People who should have their own room according to the Bedroom Standard are:

• married or cohabiting couple
• single parent
• person aged 16 years and over
• pair of same-sex persons aged 10 to 15 years
• person aged 10 to 15 years paired with a person under 10 years of the same sex
• pair of children aged under 10 years, regardless of their sex
• person aged under 16 years who cannot share a bedroom with someone in 4, 5 or 6 above

An occupancy rating of:

• -1 or less: implies that a household's accommodation has fewer rooms than required (overcrowded)
• +1 or more: implies that a household's accommodation has more rooms than required (under-occupied)
• 0: suggests that a household's accommodation has an ideal number of rooms

The number of rooms is taken from Valuation Office Agency (VOA) administrative data for the first time in 2021. The number of rooms is recorded at the address level. This means that for households that live in a shared dwelling, the available number of rooms are counted for the whole dwelling in VOA, and not each individual household.

VOA's definition of a room does not include bathrooms, toilets, halls or landings, kitchens, conservatories or utility rooms. All other rooms, for example, living rooms, studies, bedrooms, separate dining rooms and rooms that can only be used for storage are included.

“Occupancy rating for rooms” is split into 6 categories including total.

The estimates are as at Census Day, 21 March 2021.

PLEASE NOTE: This record has been retired. It has been superseded by: https://environment.data.gov.uk/dataset/f81508d3-cf5a-44ed-ae7e-452be665af84 This dataset is a product of a national assessment of flood risk for England produced using local expertise. It is produced using the Risk of Flooding from Rivers and Sea data which shows the chance of flooding from rivers and/or the sea, based on cells of 50m. Each cell is allocated one of four flood risk categories, taking into account flood defences and their condition.

This dataset uses OS address data and Royal Mail postcode data to show how many properties are in each of four flood risk categories in each postcode, based simply on the category allocated to the cell that each property is in.

The newGeoSure Insurance Product (newGIP) provides the potential insurance risk due to natural ground movement. It incorporates the combined effects of the 6 GeoSure hazards on (low-rise) buildings. This data is available as vector data, 25m gridded data or alternatively linked to a postcode database - the Derived Postcode Database. A series of GIS (Geographical Information System) maps show the most significant hazard areas. The ground movement, or subsidence, hazards included are landslides, shrink-swell clays, soluble rocks, running sands, compressible ground and collapsible deposits. The newGeoSure Insurance Product uses the individual GeoSure data layers and evaluates them using a series of processes including statistical analyses and expert elicitation techniques to create a derived product that can be used for insurance purposes such as identifying and estimating risk and susceptibility. The Derived Postcode Database (DPD) contains generalised information at a postcode level. The DPD is designed to provide a 'summary' value representing the combined effects of the GeoSure dataset across a postcode sector area. It is available as a GIS point dataset or a text (.txt) file format. The DPD contains a normalised hazard rating for each of the 6 GeoSure themes hazards (i.e. each GeoSure theme has been balanced against each other) and a combined unified hazard rating for each postcode in Great Britain. The combined hazard rating for each postcode is available as a standalone product. The Derived Postcode Database is available in a point data format or text file format. It is available in a range of GIS formats including ArcGIS (.shp), ArcInfo Coverages and MapInfo (.tab). More specialised formats may be available but may incur additional processing costs. The newGeoSure Insurance Product dataset has been created as vector data but is also available as a raster grid. This data is available in a range of GIS formats, including ArcGIS (.shp), ArcInfo coverage's and MapInfo (.tab). More specialised formats may be available but may incur additional processing costs. Data for the newGIP is provided for national coverage across Great Britain. The newGeoSure Insurance Product dataset is produced for use at 1:50 000 scale providing 50m ground resolution. This dataset has been specifically developed for the insurance of low-rise buildings. The GeoSure datasets have been developed to identify the potential hazard for low-rise buildings and those with shallow foundations of less than 2 m deep. The identification of ground instability and other geological hazards can assist regional planners; rapidly identifying areas with potential problems and aid local government offices in making development plans by helping to define land suited to different uses. Other users of these data may include developers, homeowners, solicitors, loss adjusters, the insurance industry, architects and surveyors.

Doorda's UK Vulnerability Data provides a comprehensive database of over 1.8M postcodes sourced from 30 data sources, offering unparalleled insights for location intelligence and analytics purposes.

Volume and stats: - 1.8M Postcodes - 5 Vulnerability areas covered - 1-100 Vulnerability rating

Our Residential Real Estate Data offers a multitude of use cases: - Market Analysis - Identify Vulnerable Consumers - Mitigate Lead Generation Risk - Risk Management - Location Planning

The key benefits of leveraging our Residential Real Estate Data include: - Data Accuracy - Informed Decision-Making - Competitive Advantage - Efficiency - Single Source

Covering a wide range of industries and sectors, our data empowers organisations to make informed decisions, uncover market trends, and gain a competitive edge in the UK market.

Click on the title for more details and to download the file. (File Size - 45MB)

A best-fit lookup between postcodes, frozen 2011 Census Output Areas (OA), Workplace Zones (WZ), Lower Layer Super Output Areas (LSOA), Middle Layer Super Output Areas (MSOA) and current local authority districts (LAD) along with OA, WZ, and LAD classifications as at May 2020 in the UK. Postcodes are best-fitted by plotting the location of the postcode's mean address into the areas of the output geographies. (File size 46MB).Field Names - PCD7, PCD8, PCDS, DOINTR, DOTERM, USERTYPE, OSEAST1M, OSNRTH1M, OA11CD, OAC11CD, OAC11NM, WZ11CD, WZC11CD, WZC11NM, LSOA11CD, LSOA11NM, MSOA11CD, MSOA11NM, LADCD, LADNM, LADNMW, LACCD, LACNMField Types - Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, Text, TextField Lengths - 7, 8, 8, 6, 6, 1, 6, 7, 9, 3, 48, 9, 2, 60, 9, 63, 9, 35, 9, 36, 35, 3, 48

This method returns LOAC Supergroups and Groups, including their name and descriptions. The results are determined by the inclusion of the submitted postcode/coordinates/UPRN within the corresponding LOAC Supergroup and Group.

The London Output Area Classification (LOAC) is a geodemographic that summarises the built and population characteristics of all 2021 Output Areas within Greater London. The region is organised into a hierarchical typology composed of 7 Supergroups and 16 Groups. The classification was created from the 2021 Census data.

The newGeoSure Insurance Product (newGIP) provides the potential insurance risk due to natural ground movement. It incorporates the combined effects of the 6 GeoSure hazards on (low-rise) buildings. This data is available as vector data, 25m gridded data or alternatively linked to a postcode database the Derived Postcode Database. A series of GIS (Geographical Information System) maps show the most significant hazard areas. The ground movement, or subsidence, hazards included are landslides, shrink-swell clays, soluble rocks, running sands, compressible ground and collapsible deposits. The newGeoSure Insurance Product uses the individual GeoSure data layers and evaluates them using a series of processes including statistical analyses and expert elicitation techniques to create a derived product that can be used for insurance purposes such as identifying and estimating risk and susceptibility. The Derived Postcode Database (DPD) contains generalised information at a postcode level. The DPD is designed to provide a summary value representing the combined effects of the GeoSure dataset across a postcode sector area. It is available as a GIS point dataset or a text (.txt) file format. The DPD contains a normalised hazard rating for each of the 6 GeoSure themes hazards (i.e. each GeoSure theme has been balanced against each other) and a combined unified hazard rating for each postcode in Great Britain. The combined hazard rating for each postcode is available as a standalone product. The Derived Postcode Database is available in a point data format or text file format. It is available in a range of GIS formats including ArcGIS (.shp), ArcInfo Coverages and MapInfo (.tab). More specialised formats may be available but may incur additional processing costs. The newGeoSure Insurance Product dataset has been created as vector data but is also available as a raster grid. This data is available in a range of GIS formats, including ArcGIS (.shp), ArcInfo coverages and MapInfo (.tab). More specialised formats may be available but may incur additional processing costs. Data for the newGIP is provided for national coverage across Great Britain. The newGeoSure Insurance Product dataset is produced for use at 1:50 000 scale providing 50 m ground resolution. This dataset has been specifically developed for the insurance of low-rise buildings. The GeoSure datasets have been developed to identify the potential hazard for low-rise buildings and those with shallow foundations of less than 2 m deep. The identification of ground instability and other geological hazards can assist regional planners; rapidly identifying areas with potential problems and aid local government offices in making development plans by helping to define land suited to different uses. Other users of these data may include developers, homeowners, solicitors, loss adjusters, the insurance industry, architects and surveyors. Version 7 released June 2015.

This method returns total crime rates, crime rates by crime types, area ratings by total crime, and area ratings by crime type for small areas (Lower Layer Super Output Areas, or LSOAs) and Local Authority Districts (LADs). The results are determined by the inclusion of the submitted postcode/coordinates/UPRN within the corresponding LSOA or LAD.

All figures are annual (for the last 12 months).

The crime rates are calculated per 1,000 resident population derived from the census 2021.

The dataset is updated on a monthly basis, with a 3-month lag between the current date and the most recent data.

This dataset provides detailed information on the 2019 Index of Multiple Deprivation (IMD) for Birmingham, UK. The data is available at the postcode level and includes the Lower Layer Super Output Area (LSOA) information.Data is provided at the LSOA 2011 Census geography.The decile score ranges from 1-10 with decile 1 representing the most deprived 10% of areas while decile 10 representing the least deprived 10% of areas.The IMD rank and decile score is allocated to the LSOA and all postcodes within it at the time of creation (2019).Note that some postcodes cross over LSOA boundaries. The Office for National Statistics sets boundaries for LSOAs and allocates every postcode to one LSOA only: this is the one which contains the majority of residents in that postcode area (as at 2011 Census).

The English Indices of Deprivation 2019 provide a detailed analysis of relative deprivation across small areas in England. The Crime Deprivation dataset is a key component of this index, measuring the risk of personal and material victimization at the local level. This dataset includes indicators such as recorded crimes for violence, burglary, theft, and criminal damage. It helps identify areas with high levels of crime, guiding policy interventions and resource allocation to improve safety and reduce crime rates.