For question use either UnBBayes or Netica to build your models. Note that you must provide your modeling assumptions and a running model (i.e. not just a screenshot). Like Noisy-OR BN models.SnowLair, Inc., has issued a contract to design an upgrade to an existing firewall system. Based on the results of a preliminary design review, the program manager believes that a critical driver of success of the upgrade is the impact to the system that a new, untested technology that the company is proposing to use would have. Based on engineering analyses, the program manager believes: • There is a 75% chance that the new technology will be deployed successfully in the new system without impacting cost, schedule, and performance, a 20% chance of a moderate impact, and a 5% chance of a severe impact. If there is no impact, the program manager forecasts a 75% chance of a cost overrun smaller than $10K, and a 25% chance of an overrun greater than $10K. A moderate impact means 60% chance of a cost overrun smaller than $10K, a 40% chance of overrun greater than $10K. A severe impact means a 20% chance of no overrun, a 80% chance of greater than $10K overrun. • If there is no impact, the program manager forecasts an 80% chance of on-time delivery and a 20% chance of a delay. A moderate impact means 50% chance of on- time delivery and a 50% chance of a delay. A severe impact means 5% chance of on- time delivery and a 95% chance of a delay. • If there is no impact, the program manager forecasts a 90% chance of an acceptable performance and a 10% chance of unacceptable performance. A moderate impact means 70% chance of an acceptable performance and a 30% chance of unacceptable performance. A severe impact means a 10% chance of an acceptable performance and a 90% chance of unacceptable performance. a. Construct a Bayesian network to model this problem. b. The program manager also wants to estimate the fairness of the procurement process. For this, he will only consider the procurement a failure if a cost overrun greater than $10K, a delayed delivery, and an unacceptable performance all occur *at the same time", while all other combinations of those factors would be considered a fair project. He estimates that Cost Overrun impacts Fairness of a project at 80%, Delivery Time has a 40% impact, while Performance would have a 60% impact. He also believes that the mechanisms by which Cost Overrun, Delivery Time, and Performance affect Fairness are mutually independent. Calculate what is the posterior probability of a fair procurement given that a moderate impact would happen
For question use either UnBBayes or Netica to build your models. Note that you
must provide your modeling assumptions and a running model (i.e. not just a screenshot). Like Noisy-OR BN models.SnowLair, Inc., has issued a contract to design an upgrade to an existing firewall system. Based on the results of a preliminary design review, the program manager believes that a critical driver of success of the upgrade is the impact to the system that a new, untested technology that the company is proposing to use would have. Based on engineering analyses, the program manager believes:
• There is a 75% chance that the new technology will be deployed successfully in the new system without impacting cost, schedule, and performance, a 20% chance of a moderate impact, and a 5% chance of a severe impact.
If there is no impact, the program manager forecasts a 75% chance of a cost overrun smaller than $10K, and a 25% chance of an overrun greater than $10K. A moderate impact means 60% chance of a cost overrun smaller than $10K, a 40% chance of overrun greater than $10K. A severe impact means a 20% chance of no overrun, a 80% chance of greater than $10K overrun.
• If there is no impact, the program manager forecasts an 80% chance of on-time delivery and a 20% chance of a delay. A moderate impact means 50% chance of on- time delivery and a 50% chance of a delay. A severe impact means 5% chance of on- time delivery and a 95% chance of a delay.
• If there is no impact, the program manager forecasts a 90% chance of an acceptable performance and a 10% chance of unacceptable performance. A moderate impact means 70% chance of an acceptable performance and a 30% chance of unacceptable performance. A severe impact means a 10% chance of an acceptable performance and a 90% chance of unacceptable performance.
a. Construct a Bayesian network to model this problem.
b. The program manager also wants to estimate the fairness of the procurement process. For this, he will only consider the procurement a failure if a cost overrun greater than $10K, a delayed delivery, and an unacceptable performance all occur *at the same time", while all other combinations of those factors would be considered a fair project. He estimates that Cost Overrun impacts Fairness of a project at 80%, Delivery Time has a 40% impact, while Performance would have a 60% impact. He also believes that the
Step by step
Solved in 4 steps with 1 images
