Story By:
You Li
What kind of data do you need? Elaborate this for each small story
• city of residence
• commute route (highway, freeway)
• commute time
• commute frequency per week
• road conditions ratings based on the commute route and city of residence
• the annual average cost of operating a vehicle
• the extra annual vehicle operating cost due to bad road conditions, including accelerated vehicle depreciation, additional repair costs, and increased fuel consumption and tire wear.
• Time wasted per trip due to traffic congestion on the road
• Wasted fuel due to traffic congestion (financial cost=time wasted*fuel wasted)
• Financial cost due to crashes (as a result of bad road conditions) in the forms of lost income and workplace productivity and increased insurance premium.
• the median household income by city
Why do you need the data?
1. To predict the density of using the road:
commute time
commute frequency per week
2. To predict the road condition:
city of residence
commute route
road conditions ratings based o the commute route and city of residence
3. To predict the financial impact directly and indirectly associated with congestion, accidents, and fixes:
road conditions ratings based on the commute route and city of residence
• the annual average cost of operating a vehicle
• the extra annual vehicle operating cost due to bad road conditions, including accelerated vehicle depreciation, additional repair costs, and increased fuel consumption and tire wear.
• Time wasted per trip due to traffic congestion on the road
• Wasted fuel due to traffic congestion (financial cost=time wasted*fuel wasted)
• Financial cost due to crashes (as a result of bad road conditions) in the forms of lost income and workplace productivity and increased insurance premium.
4. To compare the cost in proportion to median household income
• the median household income by city
How do you use the data in your Machine Learning applications?
• I expect the ML can calculate the estimated expense of operating a vehicle based on the road condition of the area and density of commute
• I expect the ML can calculate the estimated additional expense of operating a vehicle due to bad roads, traffic congestion, and accidents.
• I expect the ML to tell me whether I should expect to spend more (or less) on fixing the cars due to bad road conditions than the other Michigan commuters? if so, by how much?
• I expect the ML to tell me whether the budget on fixing the cars was in disappropriation to my household income as compared to other Michigan households?
How do you get the data?
Small scale: Self-reported data from commuters on expense; the road conditions data could be retrieved from SEMCOG from the previous submission
Bigger scale (context):TRIP a national NGO published a study that estimated the average cost per commuter due to bad road conditions.
Do you know any specific data set you may use for analysis of your story?
TRIP a national NGO published a study that estimated the average cost per commuter due to bad road conditions. The summary of the report is here: https://tripnet.org/wp-content/uploads/2019/07/Fact_Sheet_MI.pdf
This is the full length of the report where it has some cost estimate per commuter by city, and the highway road condition: https://tripnet.org/wp-content/uploads/2017/04/MI_Progress_and_Challenges_TRIP_Report_March_2019.pdf
Data Type for each your data – for examples, number, text, images, video, audio, social network, database
numbers, text, maybe images of the bad roads and the car damages due to bad road conditions?
What kind of analysis would you do? (Correlation, frequency, cross-tabulation, average, classification, regression, ANOVA, etc.)
average, correlation, regression, t-test
What kind of measurements would you use? (Data quality, privacy, etc.)
privacy?
What kind of visualization would be useful? (Submit your visual here)
bar graph of repair cost or cost per commuter by location on a map?
it could also be a bar graph of average cost per commuter on a level of low, moderate, high, and extremely high as compared to the state average of the cost per commuter.