Central Limit Theorem for Mean
1. In a machine learning project, you're using a dataset where the average runtime of a
specific algorithm is known to be 10 seconds with a standard deviation of 2 seconds.
You decide to run the algorithm 100 times and calculate the mean runtime. What is
the probability that the mean runtime will be less than 9.8 seconds?
Ans: the probability that the mean runtime will be less than 9.8 seconds is 0.1587 or
15.87%.
2. In a deep learning model training process, the average time taken for a batch of 64
images is 12 minutes with a standard deviation of 3 minutes. What is the probability
that in 25 training sessions, the average time per batch is more than 13 minutes?]
Ans: the probability that the average time per batch is more than 13 minutes is 0.0475
or 4.75%.
Central Limit Theorem for Sample Proportion
1. You are developing a classification model where 80% of the instances are correctly
classified. If you randomly select 150 instances, what is the probability that the
proportion of correctly classified instances in your sample is less than 75%?
Ans: The probability that the proportion of correctly classified instances is less than
75% is 0.0630 or 6.30%.
2. In an A/B testing scenario, 45% of users in group A converted, while 50% of users in
group B converted. If you randomly select 200 users from group B, what is the
probability that the sample proportion of conversions in group B is greater than 55%?
Ans: The probability that the sample proportion of conversions in group B is greater
than 55% is 0.0793 or 7.93%.
Central Limit Theorem for Sum of Observations
1. You are analysing the total prediction times for a neural network model across 30 test
instances. The time taken per instance is normally distributed with a mean of 5
seconds and a standard deviation of 1.5 seconds. What is the probability that the total
prediction time exceeds 160 seconds?
Ans : The probability that the total prediction time exceeds 160 seconds is 0.1112 or
11.12%.
2. In a recommendation system, the average time taken to compute recommendations for
a user is 2 seconds, with a standard deviation of 0.8 seconds. If you need to compute
recommendations for 50 users, what is the probability that the total computation time
will be less than 95 seconds?
Ans: The probability that the total computation time will be less than 95 seconds is
0.1894 or 18.94%.
