While I was driving in Japan one day, a country where most cars are mostly white, black, and any shade of grey in between, I have found myself wondering if the color of a car had any impact on its probably to be into an accident. Indeed I would expect a bright red car to be more noticeable and therefore be less involved in car accidents.
I unfortunately haven't found data containing the color of the cars involved in accidents. Insurance companies do apply a premium to red cars but it could be for a higher risk of theft and not of car accidents.
I have however found the very detailed records on the car accidents happening in France collected by the French interior ministry. They are so far the best I have found and I have decided to look in details at these records to get a better understanding of the factors leading to accidents. I really wonder if there are some practices that could reduce the risk of an accident. For example:
- Is it safer to walk on the side of the road facing the incoming cars or turning our backs to them ?
- Is the public lighting system sufficient or are the accidents still occuring mostly in poorly lit roads ?
- Does daylight saving help reduce or increase the number of accidents ?
- Can we predict the severity of an accident for each road user based on available information, maybe to guide emergency response teams ?
The data was obtain from the French government open data website under the Open Licence. https://www.data.gouv.fr/fr/datasets/base-de-donnees-accidents-corporels-de-la-circulation/
- Section1 - Data wrangling
- Section2 - Data visualization
- Section3 - Predict severity
- a) Preparation of features and target
- b) Machine learning
- Jupyter Notebook
- Python with various modules:
- Pandas
- Matplotlib
- Seaborn
- MariaDB database
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent