You are a machine learning specialist working for a research lab that is studying the crime rates in various cities across Europe. Your task is to project if the crime rate, or number of crimes per day, in the cities in your study group will rise in the next month. You have crime per day data for the cities in your study going back 100 years. You have calculated the mean for each city and most cities present a distribution that looks like the following when you visualize your data in a plot:
In order to create a proper repository for your machine learning model, you need to understand the distribution of your data. Based on your data source and your visualization, which type of distribution do you have?

Answer options:

A.Binomial distribution
B.Poisson distribution
C.Normal distribution
D.Gaussian distribution

Correct Answer: B
Option A is incorrect. A binomial distribution has a binary outcome, either success or failure. Your data describes the count of events (crimes) in a given period of time (per day).
Option B is correct. A poisson distribution is represented as the count of events in a given period of time. This fits your data: number of crimes per day.
Option C is incorrect. A normal distribution has data that distributes evenly on either side of the mean. A normal distribution forms a “bell-shaped curve.” Your data does not center around the mean, your data skews towards 0.
Option D is incorrect. A gaussian distribution is another name for a normal distribution.
Reference:
Please see the Towards Data Science article titled The Poisson Distribution (https://towardsdatascience.com/the-poisson-distribution-103abfddc312), the Analysis Factor article titled Differences Between the Normal and Poisson Distributions (https://www.theanalysisfactor.com/differences-between-normal-and-poisson-distributions/), the the Make Me An Analyst article titled NORMAL DISTRIBUTION, BINOMIAL DISTRIBUTION & POISSON DISTRIBUTION (http://makemeanalyst.com/normal-distribution-binomial-distribution-poisson-distribution/)