Home
Subjects
Explanations
Create
Study sets, textbooks, questions
Log in
Sign up
Upgrade to remove ads
Only $2.99/month
CHEN 460 exam 1
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (59)
sample average
average observations; estimate of unknown population mean
mean
average of population values
median
half data above 0.5 and half data below 0.5: less influenced by outliers than is sample mean
mode
highest probability of pdf, greatest frequency of pmf (discrete probability mass function)
n
sample size or number of equally weighted observations
bayes model
evaluates the probability of evidence of an event but not event itself, adaptive to casual information or evidence as becomes available to revise degree of belief in the event occurrence probability, simplest hypothesis or model that is consistent with the data therefore the most plausible
modeling
useful to avoid numerous repetitions and to enable prediction beyond observed events
subjective model
based on personal assessment or judgement by one or more experts based on experience with similar events, expert judgment (degree of belief) important source of information to supplement generic data
poisson distribution
probability of each number of relatively rare events,x, that could occur in a time period, one parameter mean value and variance, provides approximation to the binomial distribution for total number events,n, but not known
Binomial Distribution
discrete probability mass function (pmf) models the probability that an outcome will occur x times out of n trials, regardless of the order of occurrences and non-occurrences
hazard
an agent or condition with intrinsic potential to cause harm
risk
measure of potential for a hazard to result in harm or damages based on probability and consequence of an adverse event involving hazard
voluntary risk
hazards of activities without prior consent
two most sensitive variables that influence acceptable risk:
benefit and control
bayesian
combines three categories of information: generic information, expert judgement, system observations
aleatory uncertainty
due to randomness, fluctuations in behavior
expected value
weighted mean of that variable function
variance
measure distribution width
epistemic uncertainty
amount and quality of data
joint
common elements; not mutually exclusive
disjoint
no common elements ; mutually exclusive
traditional risk analysis
manage technical systems; decisions based on point value alone and assuming independence of components-easier to learn but unrealistic and therefore costly
system approach RA
manage socio-technical systems and their organizations; decisions based on ranges and distributions of information and including significant interdependencies -more realistic and cost effective
deterministic
predict nearly exact or unique result
probabilistic
using uncertainty data and information, predict output ranges and distributions of variables or of system behavior to estimate uncertain events that could result in losses
uncertainties
due to ranges of random critical variable and to parameters must be represented in models or distributions to estimate confidence intervals for validation, probability ranges, and consequence ranges under the give conditions of ourome events
Risk assessment questions:
what can go wrong? how likely? what are the consequences?
Types of Risk Assessment
Quantitative Risk Assessment
Semi-quantitative RA
Qualitative RA
Quantitative risk assessment
calculate risk in form of a numerical probability distribution of an event and the event consequence magnitude distribution, needed for most critical part of a system
semi-quantitative risk assessment
use order of magnitude for frequency and for outcome magnitude for non-critical or non-sensitive parts of a system
qualitative risk assessment
use of linguistic or ordinal scales for probability or the outcome magnitude-often used for screening but ranges must be defined for unambiguous risk designations
citizen engineer
expected to help educate stakeholders within risk governance program and to create and develop a culture of trust and confidence in risk management among stakeholders including the broad community
main question to be answered by a cumulative F-N profile:
-what is the cumulative frequency or probability of outcome events with consequence levels equal to and greater than a given threshold,c, to be reduced in probability occurrence?
-does the cumulative occurrence frequency or probability decline at an acceptable rate, based on tolerable standards, as the consequence level increases?
frequentist
data from observations of even occurrences n/N; not useful for rare events
subjectivist
info from the available knowledge about event occurrences; exerts estimate n/N values and ranges
Probability axiom 1
0<P(A)<1 for all Pr values within [0,1]
probability axiom 2
Total probability with in the sample space,S: P(S)=1
probability axiom 3
for mutually exclusive A,B, how to add Pr values: P[A or B]=P[A]+P[B] and for rare event approximation REA
probability axiom 4
fundamental rule of conditional probability or how to calculate conditional Pr, consistent with Bayes model
quality
amount by which product satisfies the requirements of the users. product quality is partly a function of design and conformance to design specifications during manufacture
reliability
concerned with how long the product continues to function once it becomes operational; attributes to quality and optimally managed engineering system
maintainability
ability to restore to working condition; contributes to quality and optimally managed engineering system
residual life
mean life or expected remaining life of a component given the probability of its survival to time T=To
conditional or residual MTTF
mean time to failure of a component given the Pr it has not failed or is working at T=To
explaining away
In BN, the information of an effect, the presence of an observed cause renders an alternative cause that is not observed less likely
point value decisions
based only on point values, such as mean, median, mode
value interval decisions
based on estimated ranges of variable values without regard to the relative probability of the values throughout the intervals
value distribution decision
based on all information including the probability of each value within a range
steps in risk assessment
hazard identification, barrier identification, barrier performance assessment, exposure assessment, risk characerization
what are the probability point value estimation approaches?
frequentist, subjective, bayesian
what are the 3 types of risk assessment models?
deterministic, probabilistic, uncertainties
risk profile
plot of risk variables, such as probabilities or frequencies along Y axis, and consequence severities along X axis
intercausal reasoning
in bayesian networks, relative credibility among 2 or more causes of common effect
diagnostic reasoning
in BN, from symptom to cause in opposite direction of arcs
predictive reasoning
In BN, finding new info about causes to beliefs about effects or symptoms, in direction of arcs
exponential distribution (CFR)
1 parameter, continuous distribution models the time of the occurrence of an event
poisson distribution (CFR)
1 parameter, p, models number event occurrences with in time t
F-N curve
display frequency of fatalities and number of occurrences
Risk matrix
represents the results of a scenario outcome event risk assessment levels, identify the categories of risk levels: acceptable, unacceptable, conditionally acceptable
Sets with similar terms
Reading 7: Risk Measurement and Risk Att…
66 terms
Chapter 17 Definitions
31 terms
291 information test 2
12 terms
SS 2
35 terms
Other sets by this creator
ENGR 482 Exam 2
57 terms
Bioprocess Exam 3
43 terms
SENG 310 Exam 3
126 terms
Bioprocess Exam 2
101 terms