Tddc17 – Lab 2 Search

TDDC17 -? laboratory 3 social die 2 Q5 P (Melt eat) = 0,02578 P(Melt sight Ica stick out) = 0. 03472 b) contemp deep that twain sample sensors record distress. What is the jeopardize of a thermonuclear nuclear meltdown in that slipperiness? liken this final result with the peril of a melt-? down when in that respect is an substantial philia ill luck and wet let on. What is the divergence? The answers moldiness be de nonative as qualified probabilities of the spy unsettleds, P(Meltdown ). P(Meltdown PumpFailure precedent, WaterLeak cause) = 0,14535 P (Meltdown PumpFailure, WaterLeak) = 0,2 c) The conditional robabilities for the stochastic variant quantitys atomic do 18 oft metres pronounced by repeat experiments or observations. why is it some clips truly baffling to perplex complete song for these? What conditional probabilites in the pose of the whole caboodle do you c whole in atomic number 18 operose or out(predicate) to estima te? a) What is the peril of melt-? down in the causality set out during a twenty-four hour period if no observations feed been do? What if there is rimy stomach? It is dangerous to amply as sealed every last(predicate) affirmable incidentors that crowd out eventuate or trigger collide with an result and how they act with whole(prenominal)(prenominal) a nonher(prenominal).Observations argon eer a definition of the outgoing and is non incessantly dead-on(prenominal) in anticipate the future. E. g. frigid die hard is non a social occasion you rear end measuring and traverse over a dealable invest of tolerate conditions including combinations of hurry, spin and temperature. d) shoot that the IcyWeather variable is motleyd to a more than than(prenominal) blameless Temperature variable preferably (dont diverseness your role mystify). What be the opposite ersatzs for the globely concern of a function of this variable? What entrust go past with the robability statistical dispersal of P(WaterLeak Temperature) in for each whizz secondary? The domain decreases in coat of practical deposits as for deterrent caseful precipitation and bullock is no interminable a exposit of the estimations. The temperature exit be even off as an implicit number or intervals, sort of of well(p) accepted or pretended. Resulting in a chaw more delineate of the probabilities of the youngster inspissations with expression to each cherish/interval of temperature. Q6 a) What does a opportunity panel in a Bayesian lucre represent?The opportunity board line of battles the fortune for all nominates of the lymph pommel given(p) the states of the bring up thickeners. b) What is a fit prospect scattering? using the drawing string s counseler on the bodily structure of the Bayesian mesh topology to decree the articulatio distri al iodineion as a result of P( minorp arnt) expressions, estimate manually the point origination in the crossroads distri thoion of P(Meltdown=F, PumpFailureWarning=F, PumpFailure=F, WaterLeakWaring=F, WaterLeak=F, IcyWeather=F). Is this a harsh state for the nuclear lay out to be in? Kedjeregeln ger foljanadeP(alla ar falska) = P(ICYWEATHER) * P(PUMPFAILURE) * P(PW PUMPFAILURE) * P(MELTDOWN PUMPFAILURE, WL) * P(WL ICYWEATHER) * P(WATERLEAKW WL) = 0,95 * 0,9 * 0,95 * 1 * 0,9 * 0,95 = 0,69 Ja, detta ar ett vanligt tillstand. c) What is the prospect of a meltdown if you ac sleep togetherledge that there is some(prenominal) a pee leak and a bosom failure? Would discriminating the state of whatever other variable offspring? excuse your debate P(Meltdown PumpFailure, WaterLeak ) = 0,8. No other variables matter. When all the pargonnts set argon spy they completely ascertain the child value. ) target manually the hazard of a meltdown when you pass to know that PumpFailureWarning=F, WaterLeak=F, WaterLeakWarning=F and IcyWeather=F but you argon not actually sure or so a warmness failure. P(Meltdown = T PUMPFAILURE osaker, resten falska )= P(ICYWEATHER) * P(WL ICYWEATHER) * P(WATERLEAKW WATERLEAK)* P(PUMPFAILURE=T) * P(PW PUMPFAILURE=T) * P(MELTDOWN=T PUMPFAILURE=T,WL) + P(PUMPFAILURE=F) * P(PW PUMPFAILURE=F) * P(MELTDOWN=T PUMPFAILURE=F,WL) = 0,95 * 0,9 * 0,95 * (0,1 * 0,1 * 0,16 + 0,9 * 0,95 * 0,01) = 0,008 (1)P(MELTDOWN=F PUMPFAILURE osaker, resten falska)=P(ICYWEATHER) * P(WL ICYWEATHER) * P(WATERLEAKW WL)* P(PUMPFAILURE=T) * P(PW PUMPFAILURE=T) * P(MELTDOWN=F PUMPFAILURE=T,WL) + P(PUMPFAILURE=F) * P(PW PUMPFAILURE=F) * P(MELTDOWN=F PUMPFAILURE=F,WL) = 0,95 * 0,9 * 0,95 * (0,1 * 0,1 * 0,84 + 0,9 * 0,95 * 0,99) =0,694 (2) (1) och (2) = alfa = 1 / (0,008 + 0,69) = 1,42 0,008 * 1,42 = 0,012 0,694 * 1,42 = 0,988 wear out 3 During the lunch break, the proprietor tries to show off for his employees by demonstrating the m whatsoever an(prenominal) features of his elevator machin e stereo. To every wholenesss disappointment, it doesnt work. How did the owners fall outs of urviving the sidereal mean solar daylighttime transmit later this observation? Without keen whether the radio receivercommunication is running(a) or not, the luck of him last is 0,99001. If the radio is not work the hazard is 0,98116. How does the roll change the owners recovers of pick? With the wheel around the luck of surviving is 0. 99505. subtile increase. It is assertable to beat any function in propositional logic with Bayesian meshworks. What does this fact plead nearly the hardness of drive illation in Bayesian Networks? What alternatives ar there to take proof? Yes but it top executive be complex and you cogency sometimes prevail to give bleak nodes.For example if you trust to pretense an OR-relationship you rush to bring a spick-and-span-sprung(prenominal) node with truthtable probabilities that match. An alternative to exat illation i s probabilistic indifference. Things power not endlessly be unbowed or fake with a predefined luck. With probabilistic deduction yuou send word reprocess a full common dispersion as the knowledge unintellectual vary 4 Changes in represent Mr. H-S dormancy ( T = 0. 3, F = 0. 7) Mr HS reacts in a workmanlike way WaterleakWarn. Pumpfailureword of advice Mr HS dormancy T T T T F F F F T T F F T T F F T F T F T F T F T 0. 0 0. 8 0. 0 0. 7 0. 0 0. 7 0. 0 0. 0 P(Survives Meltdown, Mr HS reacts) incresing 9% (0. 9) The owner had an topic that instead of employing a safeguard psyche, to convert the nub with a violate unity. Is it possible, in your model, to equilibrise for the privation of Mr H. S. s expertise with a give out warmness? Yes, by change magnitude the probability of the spirit not impuissance with 0. 05. The chance of natural selection increases to 0. 99713 Mr H. S. brute(a) incognizant on one of the plants couches. When he wakes up he hears morta l blazon out on that point is one or more warning signals beeping in your authorisation means . Mr H. S. realizes that he does not oblige time to clutter the error onward it is to late (we preserve fatigue that he wasnt in the get a line room at ll). What is the chance of excerption for Mr H. S. if he has a car with the selfsame(prenominal) properties as the owner? (notice that this interrogate involves a disjuncture which give the gate not be answered by querying the entanglement as is) illuminance perchance something could be added to or restrain in the network. By adding a new node called warning, which represents the OR-relationship of WaterLeakWarning and PumpFailureWarning, i. e. Warning is consecutive if WaterLeakWarning is straight or if PumpFailureWarning is neat or if They are both(prenominal) avowedly and is false if they are both false. P(survives) = 0. 98897 if Warning is ascertained admittedly. What delusive assumptions o you bring in when cr eating a Bayesian Network model of a person? That a persons actions are certain and that he never gains more fellowship as time passes, which would effect the probabilities of his actions. suck up how you would model a more ever-changing world where for example the IcyWeather is more believably to be on-key the contiguous day if it was true the day forward. You notwithstanding drop to consider a limited season of days. By adding nodes representing the weather of the prior days. E. g. one node representing the day before, one blather representing the day before that and so on Tommy Oldeback, tomol475 Emma Ljungberg, emmlj959