Goal:
To create a classification model that can predict the severity of the traffic accident based on various variables
- Planning
- Acquire
- Prepare
- Explore
- Model
- Evaluate
- Conclusion
| Attribute | Description | |
|---|---|---|
| 1 | ID | This is a unique identifier of the accident record. |
| 2 | Source | Indicates source of the accident report (i.e. the API which reported the accident.). |
| 3 | TMC | A traffic accident may have a Traffic Message Channel (TMC) code which provides more detailed description of the event. |
| 4 | Severity | Shows the severity of the accident, a number between 1 and 4, where 1 indicates the least impact on traffic (i.e., short delay as a result of the accident) and 4 indicates a significant impact on traffic (i.e., long delay). |
| 5 | Start_Time | Shows start time of the accident in local time zone. |
| 6 | End_Time | Shows end time of the accident in local time zone. End time here refers to when the impact of accident on traffic flow was dismissed. |
| 7 | Start_Lat | Shows latitude in GPS coordinate of the start point. |
| 8 | Start_Lng | Shows longitude in GPS coordinate of the start point. |
| 9 | End_Lat | Shows latitude in GPS coordinate of the end point. |
| 10 | End_Lng | Shows longitude in GPS coordinate of the end point. |
| 11 | Distance(mi) | The length of the road extent affected by the accident. |
| 12 | Description | Shows natural language description of the accident. |
| 13 | Number | Shows the street number in address field. |
| 14 | Street | Shows the street name in address field. |
| 15 | Side | Shows the relative side of the street (Right/Left) in address field. |
| 16 | City | Shows the city in address field. |
| 17 | County | Shows the county in address field. |
| 18 | State | Shows the state in address field. |
| 19 | Zipcode | Shows the zipcode in address field. |
| 20 | Country | Shows the country in address field. |
| 21 | Timezone | Shows timezone based on the location of the accident (eastern, central, etc.). |
| 22 | Airport_Code | Denotes an airport-based weather station which is the closest one to location of the accident. |
| 23 | Weather_Timestamp | Shows the time-stamp of weather observation record (in local time). |
| 24 | Temperature(F) | Shows the temperature (in Fahrenheit). |
| 25 | Wind_Chill(F) | Shows the wind chill (in Fahrenheit). |
| 26 | Humidity(%) | Shows the humidity (in percentage). |
| 27 | Pressure(in) | Shows the air pressure (in inches). |
| 28 | Visibility(mi) | Shows visibility (in miles). |
| 29 | Wind_Direction | Shows wind direction. |
| 30 | Wind_Speed(mph) | Shows wind speed (in miles per hour). |
| 31 | Precipitation(in) | Shows precipitation amount in inches, if there is any. |
| 32 | Weather_Condition | Shows the weather condition (rain, snow, thunderstorm, fog, etc.) |
| 33 | Amenity | A POI annotation which indicates presence of amenity in a nearby location. |
| 34 | Bump | A POI annotation which indicates presence of speed bump or hump in a nearby location. |
| 35 | Crossing | A POI annotation which indicates presence of crossing in a nearby location. |
| 36 | Give_Way | A POI annotation which indicates presence of give_way in a nearby location. |
| 37 | Junction | A POI annotation which indicates presence of junction in a nearby location. |
| 38 | No_Exit | A POI annotation which indicates presence of no_exit in a nearby location. |
| 39 | Railway | A POI annotation which indicates presence of railway in a nearby location. |
| 40 | Roundabout | A POI annotation which indicates presence of roundabout in a nearby location. |
| 41 | Station | A POI annotation which indicates presence of station in a nearby location. |
| 42 | Stop | A POI annotation which indicates presence of stop in a nearby location. |
| 43 | Traffic_Calming | A POI annotation which indicates presence of traffic_calming in a nearby location. |
| 44 | Traffic_Signal | A POI annotation which indicates presence of traffic_signal in a nearby location. |
| 45 | Turning_Loop | A POI annotation which indicates presence of turning_loop in a nearby location. |
| 46 | Sunrise_Sunset | Shows the period of day (i.e. day or night) based on sunrise/sunset. |
| 47 | Civil_Twilight | Shows the period of day (i.e. day or night) based on civil twilight. |
| 48 | Nautical_Twilight | Shows the period of day (i.e. day or night) based on nautical twilight. |
| 49 | Astronomical_Twilight | Shows the period of day (i.e. day or night) based on astronomical twilight. |
- Look for project ideas
- Find the data for the project
- Explore data to identify the target variable
- Outline steps to proceed forward
- Acquired data from Kaggle
- Initially had 3 million observations
- Filtered the data down by state
- Chose only to work with California data
- Saved the copy of the data on the local machine
- Prepared a acquire.py module
- Read the data
- Dropped redundant columns
- Normalized the column names
- Dropped na
- Extracted features like the day, month, and duration
- Data conversion from boolean type to int type
- Generated dummy variables for some of the categorical variables
- Prepared a prepare.py module to store all of the functions
- Explored relationship between independent and dependent variables
- Density plot of all of the numeric variables
- Count plot of the categorical varibales
- Plotted lat and long after filtering the data by different class of target var to identify the hotspots
- Selected 15 features using selectKBest
- Used various classification algorithms to create models
- Trained these models on train data set
- Models were evaluated using their accuracy
- RandomForest performed consistenly better than other models in all train, validate, and test data
- RandomFores is the final model
- Do the entire project again but using all of the 3 million observations
- Explore if KNN can be used without having to wait 10+ hours to train the model
- Explore other classification algorithms to see if they can imporve accuracy