Fire Department Turns to BI Analytics. New York City has nearly one million buil
ID: 370991 • Letter: F
Question
Fire Department Turns to BI Analytics. New York City has nearly one million buildings, and each year, more than 3000 of them experience a major fire. The Fire Department of the City of New York (FDNY) is adding BI analytics to its arsenal of firefighting equipment. It has created a database of over 60 different factors (e.g., building location, age of the building, whether it has electrical issues, the number and location of sprinklers) in an attempt to determine which buildings are more likely to have a fire than others. The values of these parameters for each building are fed into a BI analytics system that assigns each of the city's 330,000 inspectable buildings a risk score. (FDNY doesn't inspect single and two-family homes.) Building inspectors then use these risk scores to prioritize which buildings to visit on their weekly inspections. The FDNY has roughly 350 inspectors who are trained and certified to perform their duties.
Which set of three parameters all provide measures useful in determining which buildings are more likely to have a fire than others?
a. Year the building was constructed, number of building occupants, and primary materials used in construction of the building
b. Primary materials used in the construction of the building, assessed value for property taxes, and distance from the nearest fire station
c. Distance from the nearest fire hydrant, whether or not the building has an elevator, and the number of stories in the building
d. The amount the building is insured for, distance from the nearest fire hydrant, and primary building materials used in construction of the building
Which one of the following issues is the most significant barrier to the expansion of and further investment in this program?
a. Buildings are constantly being torn down and new ones constructed. Existing buildings are constantly being renovated and remodeled. Thus, the likelihood of an individual building catching fire can vary greatly over time, without the FDNY being aware of the change in its status. The BI analytics system simply cannot handle this variation.
b. FDNY has not yet established a self-service analytics program, so individual fire inspectors are not able to view and analyze the building data.
c. Fire inspectors may not consistently evaluate a building using the key parameters of the system. As a result, a building may get a good score on one inspection and a bad score on its next inspection—even if nothing about the building has changed. The quality of the input data for the various parameters that describe a building is just too inconsistent.
d. Demonstrating that a BI analytics system was the reason behind a decrease in fires would be difficult because it involves proving that a decrease in the number of fires can be attributed to the use of the analytics system rather than other factors.
Which BI tool or technique would be most useful in predicting the likelihood of a fire in a building?
a. Data mining
b. Simple linear regression
c. Drill-down analysis on the building data
d. Dashboards
a. Year the building was constructed, number of building occupants, and primary materials used in construction of the building
b. Primary materials used in the construction of the building, assessed value for property taxes, and distance from the nearest fire station
c. Distance from the nearest fire hydrant, whether or not the building has an elevator, and the number of stories in the building
d. The amount the building is insured for, distance from the nearest fire hydrant, and primary building materials used in construction of the building
Explanation / Answer
1) Which set of three parameters all provide measures useful in determining which buildings are more likely to have a fire than others?
Solution: Option A i.e. Year the building was constructed, number of building occupants, and primary materials used in construction of the building makes a total sense in understanding the need for FDNY to visit them during their inspection. SInce these factors mentioned above are very important to undestand whether the fire will burst out or not hence it should help FDNY to act accordingly.
2) Which one of the following issues is the most significant barrier to the expansion of and further investment in this program?
Soluton: Option C i.e. Fire inspectors may not consistently evaluate a building using the key parameters of the system. As a result, a building may get a good score on one inspection and a bad score on its next inspection—even if nothing about the building has changed. The quality of the input data for the various parameters that describe a building is just too inconsistent.
It is said in data analytics terms that Garbage in is Garbage out means if you are feeding software with wrong data then it will give wrong output hence it makes no sense in investing in BI tools if they are not used effectively,
3) Which BI tool or technique would be most useful in predicting the likelihood of a fire in a building?
Solution: Though Data mining and Drill down analysis both can be used for this purpose but Drill down analysis on building data seems to be most relevent to understand the data granularity and hence FDNY should invest in it.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.