Parcial 2 Flashcards
By what is affected de actual duration
Individual Skills and Learning Curves
What does CPM do
Assumes time estimates are deterministic
Obtain task duration from previous projects
How the Pert Durations are
PERT = CPM + Probabilistic Task Times
Better por R&D type projects
Limited previous data to estimate time durations
Captures Schedule Risk
What is Beta Distribution
All values are enclosed within interval
• As classes get finer - arrive at β-
distribution
• Statistical distribution
Which is te formula for Standard Deviation
S= (b-a)/6
V=s2
Advantages for PERT
Includes Variance
Assessment of probability of Achieving a Goal
Disavantages for PERT
Data intensive -Very Time Consuming • Validity of Beta Distribution for Activity Durations • Only one Critical Path considered • Assumes independence between activity durations
What does the CPM Judgement do
Focuses attention on a subset of critical tasks
– Determine effect of shortening/lengthening tasks
– Evaluate costs of a ”crash” program
What doesnt the CPM Judgement do
Doesn’t capture task iterations, in fact …
– Prohibits iterations = “cycle error”
– Treats task durations as deterministic
What is the thing that CPM and PERT has in common
Dont allow Task iterations
Give a summary of CPM
CPM is useful, despite criticism, to identify the
critical path - focus on a subset of the project
• Slack (TS and FS) is precious
– apply flexibility to smooth resource/schedules
Give a summary of PERT
PERT treats task times as probabilistic
– Individual task durations are β-distributed
– Sums of multiple tasks are normal z-distributed
• Selective “crashing” of critical tasks can
reduce total project cost
What are the GERT Basics
Builds on Activity-on-Arrow diagrams • Discrete semantics Transactions secure resources for processing – Must wait until needed resources are available – Probabilistic Durations – Branching & Looping • Time no longer linearly increasing to right!
Which are the components of Stochastic Networks
Directed branches – Arcs – Edges – Transmittances • Directed branch – From one node (from which it emanates) – To another node (where it terminates)
Two probabilities associated with each branch: 1. The probability a branch is taken, pa, given that the node from which emanated is realized, and 2. A time, ta, required if the branch is taken (if not then the duration is 0 (zero))
You have to do GERT until you get
- The probability of realizing an output node
- The Moment Generating Function of time
to realize the output node
