1.
In a cybersecurity lab, a researcher is analyzing the relationship between CPU usage anomalies
and security breaches over 10-hour monitoring windows. Two binary events are recorded for each
window:
● Event A: A security breach occurred during the 10-hour period.
● Event B: The CPU usage spiked above a critical threshold at least once during the
10-hour period.
The researcher monitored 10,000 distinct time windows, and the observations are summarized in
the table below:
Security Breach No Breach Detected Total
Detected
CPU Usage Spiked 2000 4500 6500
CPU Usage Did Not Spike 500 3000 3500
Total 2500 7500 10000
a. What is the conditional probability that a security breach occurred, given that the CPU
usage spiked?
b. What is the conditional probability that a CPU spike occurred, given that a security
breach was detected?
c. What is the overall probability that a randomly selected monitoring window resulted in a
security breach?
d. What is the overall probability that a randomly selected monitoring window had a CPU
usage spike?
2. A factory has three machines producing bolts with the following characteristics:
○ Machine A: Produces 25% of the bolts; 5% of these bolts are defective.
○ Machine B: Produces 35% of the bolts; 4% of these bolts are defective.
○ Machine C: Produces 40% of the bolts; 2% of these bolts are defective.
A bolt is selected at random and found to be defective. What is the probability that this defective
bolt was produced by Machine B?
3. Suppose there are two classes of fruits: Apples (Class A1) and Oranges (Class A2). A fruit is
classified based on a feature X — its weight in grams.
● The true weight distributions are modeled as normal distributions but the parameters
(mean and variance) are unknown.
● You have a sample dataset of fruits with actual classes and weights (training data).
● Using this:
○ Estimate (via MLE) the parameters of weight distribution for each class (mean
and variance)
○ Use the estimated distributions to compute likelihoods (P(X|Ai)
○ Apply Bayes decision theory to classify a new fruit with weight X=150, by
finding posterior probabilities and choosing the class with the higher posterior.
Fruit class Weight (grams)
Apples (A1) 140, 145, 150, 155, 160
Orange(A2) 170, 175, 180, 185, 190
