Controlling the False Discovery Rate in Bayesian Network Structure Learning
|
|
- Roberta Simpson
- 6 years ago
- Views:
Transcription
1 False Discovery Rate in Bayesian Network Structure Learning Ioannis Tsamardinos Asnt Prof., CSD, Univ. of Crete ICS, FORTH Laura E. Brown DBMI, Vanderbilt Univ. Sofia Triantafylloy CSD, Univ. of Crete ICS, FORTH May 29, 2008 in BN Learning Ioannis Tsamardinos, Univ. of Crete 1 / 78
2 in BN Learning Ioannis Tsamardinos, Univ. of Crete 2 / 78
3 State-of-the-art BN Learning Algorithms exist that scale up to problems with thousands of variables [7] Decent quality of learning in BN Learning Ioannis Tsamardinos, Univ. of Crete 3 / 78
4 Theory of Learning-Quality Asymptotic (to sample size) proofs of consistency Under the Faithfulness assumption (e.g., PC [6], GES [2]) Rate of convergence Under specific distributional assumptions such as normality [4] in BN Learning Ioannis Tsamardinos, Univ. of Crete 4 / 78
5 Quality Assessment Empirical evaluation of the networks based on Learn Multiple Networks and estimate how reliably a feature (e.g., edge) appears [3] Relatively inefficient, does not scale up Each sampling requires learning the complete network Hinders practical applications in BN Learning Ioannis Tsamardinos, Univ. of Crete 5 / 78
6 Lack of Statistical Interpretation of Output Lack of efficiently computable, quality-estimating methods deters practitioners from using BN learning technology in BN Learning Ioannis Tsamardinos, Univ. of Crete 6 / 78
7 Local Structure Learning Problem N(T) the set of neighbors of variable T in any network that faithfully captures the data Well defined problem: N(T) is the same in all networks faithful to the data distribution in BN Learning Ioannis Tsamardinos, Univ. of Crete 7 / 78
8 Importance of inducing N(T) Variable Selection N(T) is part of the Markov Blanket of T Bayesian Network Structure Induction It can be used to learn the complete skeleton on of an Region of Interest Causal Discovery N(T) includes the Direct Causes and Direct Effects of T in BN Learning Ioannis Tsamardinos, Univ. of Crete 8 / 78
9 Problem Definition Produce an estimate N(T, a) of N(T) from finite data with a False Discovery Rate less than a given threshold α FDR = E(V/R) V the number of variables in N(T,a) not in N(T) R the number of variables in N(T,a) expected (on average) percentage of false positives in the output expected (on average) percentage of false positives in the output in BN Learning Ioannis Tsamardinos, Univ. of Crete 9 / 78
10 in BN Learning Ioannis Tsamardinos, Univ. of Crete 10 / 78
11 Certificate of exclusion All variables V = {V i } I(V i ; T Z k ) : V i is independent of T given Z k A Z s.t. I(V i ; T Z) is called a certificate of exclusion of V i from N(T) C(V i, T,S) Z S\{V i, T }, s.t.,i(v i ; T Z) True if there is a certificate of exclusion of V i from T within S Requires 2 S tests of independence to determine truth value in the worst case in BN Learning Ioannis Tsamardinos, Univ. of Crete 11 / 78
12 Learning N(T) Theorem 1. [6] If P(V ) is faithful, then V i / N(T) C(V i, T,V) Theorem 2. [8] If P(V ) is faithful, then V i / N(T) C(V i, T, E (T)) or C(V i, T, E (V i)) where E (T) N(T) and E (V i ) N(V i ) in BN Learning Ioannis Tsamardinos, Univ. of Crete 12 / 78
13 Extended Neighbor Sets E(T) any subset of V s.t., 1. is a superset of N(T) 2. V i E(T), C(V i,t,e(t)), i.e., for all variables in E(T), there is no certificate of exclusion within E(T) Contains N(T) plus the variables that cannot be d-separated from T conditioned on some subset of N(T) in BN Learning Ioannis Tsamardinos, Univ. of Crete 13 / 78
14 Learning E(T) Algorithm 1 Algorithm-E(T) 1: procedure E(T ) 2: Initialize E(T) V \T,R = V \(E(T) T) 3: repeat (arbitrarily alternate between (4) and (5)) 4: Select and move a variable in R into E(T) 5: if C(V i,t,e(t)) for some V i E(T) then 6: Remove V i from E(T) 7: end if 8: until E(T) is stable and R = 9: return E(T) 10: end procedure in BN Learning Ioannis Tsamardinos, Univ. of Crete 14 / 78
15 Learning E(T) E(T) any subset of V set s.t., 1. is a superset of N(T) 2. V i E(T), C(V i,t,e(t)) Algorithm-E(T) will return an E(T) All variables are considered and will enter E(T) eventually A variable is removed only if a certificate of exclusion is found, thus no variable in N(T) will be removed once it enters E(T) (superset of N(T)) Condition V i E(T), C(V i,t,e(t)) is directly enforced in BN Learning Ioannis Tsamardinos, Univ. of Crete 15 / 78
16 Learning E(T) Algorithm-E(T) time complexity assuming a near-perfect heuristic that will select the N(T) first: V 2 E(T), tests of independence for sparse networks quite effective heuristics exist in BN Learning Ioannis Tsamardinos, Univ. of Crete 16 / 78
17 Learning N(T) Theorem 2: V i / N(T) C(V i,t,e (T)) or C(V i,t,e (V i)) Algorithm 2 Algorithm-N(T) 1: procedure N(T ) 2: E(T) = Algorithm-E(T) 3: N(T) = E(T) neighbors plus false positives 4: for all V i E(T) do 5: E(V i ) = Algorithm-E(V i ) 6: if T / E(V i ) then 7: remove V i from N(T) 8: end if 9: end for 10: return N(T) 11: end procedure Complexity of N(T) : V E(T) 2 E(T) assuming E(T) E(V i ) in BN Learning Ioannis Tsamardinos, Univ. of Crete 17 / 78
18 Instantiations of the General Template Max-Min Parents and Children Start with E(T) = Selection heuristic: the variable in R that maximizes the minimum association with T, conditioned on all subsets of the current E(T) HITON Start with E(T) = Selection heuristic: the variable in R that maximizes pairwise association with T Slightly different strategy for searching for certificates of exclusion in BN Learning Ioannis Tsamardinos, Univ. of Crete 18 / 78
19 Asymptotic Correctness Algorithm-N(T) is correct Under faithfulness We can determine whether I(X;T Z) from data exactly in BN Learning Ioannis Tsamardinos, Univ. of Crete 19 / 78
20 of the Algorithms in BN Learning Ioannis Tsamardinos, Univ. of Crete 20 / 78
21 Practical Implementation T(X; Y Z) denote testing the respective independence T(X; Y Z) are typically implemented using statistical hypotheses tests of conditional independence Primitive Hypothesis: H i,j,k : I(V i ; V j Z k ) Let p i,j,k be the p-value returned by the test T(V i ; V j Z k ) for hypothesis H i,j,k in BN Learning Ioannis Tsamardinos, Univ. of Crete 21 / 78
22 Practical Implementation Thresholding on p i,j,k to reject or accept independencies We assume I(V i ;T Z k ) p i,t,k > t Define finite version of C: C(V i, T, V, t) Z V \{V i, T }, s.t., p i,t,k > t in BN Learning Ioannis Tsamardinos, Univ. of Crete 22 / 78
23 Finite sample: Algorithm E(T,t) Algorithm 3 Algorithm-E(T, t) 1: procedure E(T, t) 2: Initialize E(T) V \{T }, R = V \(E(T) {T }) 3: repeat (arbitrarily alternate between (4) and (5)) 4: Select and move a variable in R into E(T) 5: if C(V i, T, E(T), t) for some V i E(T) then 6: remove V i from E(T) 7: end if 8: until E(T) is stable and R = 9: return E(T) 10: end procedure Call the output E(T, t) to show dependency on t in BN Learning Ioannis Tsamardinos, Univ. of Crete 23 / 78
24 Finite Sample:Algorithm-N(T, t) Algorithm 4 Algorithm-N(T) 1: procedure N(T ) 2: E(T) = Algorithm-E(T, t) 3: N(T) = E(T) neighbors plus false positives 4: for all V i E(T) do 5: E(V i ) = Algorithm-E(V i, t) 6: if T / E(V i ) then 7: remove V i from N(T) 8: end if 9: end for 10: return N(T) 11: end procedure Call the output N(T, t) to show dependence on t in BN Learning Ioannis Tsamardinos, Univ. of Crete 24 / 78
25 Finite Sample Learning Quality What is the False Discovery Rate of Algorithm-N(T, t)? That is, what is the average proportion of false positives within the returned N(T, t)? How can we control the FDR? in BN Learning Ioannis Tsamardinos, Univ. of Crete 25 / 78
26 Statistical Error in BN Learning Ioannis Tsamardinos, Univ. of Crete 26 / 78
27 A Strategy of Controlling Error in Structure Learning 1. Express a BN-learning task as a multiple hypotheses testing problem 2. Approximate or bound the p-value of the hypotheses 3. Use statistical methods for controlling the error in BN Learning Ioannis Tsamardinos, Univ. of Crete 27 / 78
28 N(T) as a Multiple Hypotheses Testing Problem Define complex null hypotheses: H i,t : V i / N(T), equivalently (Theorem 2) H i,t : C(V i,t,e(t)) or C(V i,t,e(v i )) Rejecting H i,t implies we accept V i as a member of N(T) A False Discovery occurs when H i,t is rejected, but V i / N(T) in BN Learning Ioannis Tsamardinos, Univ. of Crete 28 / 78
29 A Strategy of Controlling Error in Structure Learning 1. Express a BN-learning task as a multiple hypotheses testing problem 2. Approximate or bound the p-value of the hypotheses 3. Use statistical methods for controlling the error in BN Learning Ioannis Tsamardinos, Univ. of Crete 29 / 78
30 Bound the p-value of H i,t p i,t is hard to compute exactly... But, can be bounded from above Theorem 3. [8] Let p i,t be the p-value of H i,t. Then, where p i,t p i,t, p i,t = max{p i,t,k, where Z k E(T) or Z k E(V i )} in BN Learning Ioannis Tsamardinos, Univ. of Crete 30 / 78
31 Bound the p-value of H i,t The p-value of a complex hypothesis H i,t is bounded by the maximum p-value of the primitive hypotheses H i,t,k : I(V i, T Z k ) in BN Learning Ioannis Tsamardinos, Univ. of Crete 31 / 78
32 A Strategy of Controlling Error in Structure Learning 1. Express a BN-learning task as a multiple hypotheses testing problem 2. Approximate or bound the p-value of the hypotheses 3. Use statistical methods for controlling the error in BN Learning Ioannis Tsamardinos, Univ. of Crete 32 / 78
33 Control FDR FDR(d): the FDR we obtain when we reject all hypotheses H i,t, such that p i,t d in BN Learning Ioannis Tsamardinos, Univ. of Crete 33 / 78
34 Control FDR Control FDR problem: Given P = {p i,t } and a desired maximum FDR level a, select as large as possible threshold d, s.t. FDR(d) a Call such a procedure F(P,a) Inverse: given threshold d, find a small as possible FDR level a guaranteed Call such a procedure F 1 (P,d) Such procedures exist, e.g., [1] in BN Learning Ioannis Tsamardinos, Univ. of Crete 34 / 78
35 Reality Creeps In H i,t : C(V i, T, E(T)) or C(V i, T, E(V i )) How do we obtain E(T) and E(V i ) to produce p i,t? Use Algorithm-E(T, t) to obtain an approximation If E(T,t) is a superset of E(T), H i,t are still tested correctly (Theorem 2) We just need to assume that we have enough power to detect dependencies at the t level for all members of N(T) in BN Learning Ioannis Tsamardinos, Univ. of Crete 35 / 78
36 Assumption About Statistical Power Assumption 1: the network is t-faithful(t ), i.e., V i N(T), then Z k V \ {T, V i }, p i,t,k < t A form of (local) Faithfulless assumption for finite sample We call t the power threshold in BN Learning Ioannis Tsamardinos, Univ. of Crete 36 / 78
37 C-FDR(T, t, a): Control FDR Control the FDR of estimating N(T) at the a-level, assuming the network is t-faithful(t ) Algorithm 5 C-FDR(T, t, a) 1: procedure C-FDR(T, t, a) 2: N(T) = Algorithm-N(T, t) 3: for all V i N(T) do 4: E(V i )= Algorithm-E(V i, t) 5: p i,t = max{p i,t,k,z k N(T) or Z k E(V i )} 6: end for 7: d = F({p i,t }, a) call d the cut-off threshold 8: return {V i : p i,t d} 9: end procedure Call the output N(T,t,a) to show dependence on t and a in BN Learning Ioannis Tsamardinos, Univ. of Crete 37 / 78
38 C-FDR(T, t, a) Remarks We only calculate the p-value bounds p i,t for V i N(T, t) We assume the network is t-faithful(t ) and so all N(T) N(T, t) no need to consider the other variables in BN Learning Ioannis Tsamardinos, Univ. of Crete 38 / 78
39 B-FDR(T, t, d): Bound FDR Bound the FDR of estimating N(T) assuming the network is t-faithful(t ) and rejecting a complex hypothesis when p i,t d Algorithm 6 B-FDR(T, t, d) 1: procedure B-FDR(T, t, d) 2: N(T) = Algorithm-N(T,t) 3: for all V i N(T) do 4: E(V i )= Algorithm-E(V i,t) 5: p i,t = max{p i,t,k,z k N(T) or Z k E(V i )} 6: end for 7: a = F 1 ({p i,t },d) 8: return a 9: end procedure in BN Learning Ioannis Tsamardinos, Univ. of Crete 39 / 78
40 N(T) FDR Procedures C-FDR(T, t, a) can be used to learn an estimate of N(T) with a maximum predefined FDR level a B-FDR(T, t, d) can be used to bound the FDR level of an algorithm returning N(T) using a cut-off threshold d in BN Learning Ioannis Tsamardinos, Univ. of Crete 40 / 78
41 N(T) FDR Procedures Time complexity: same as running Algorithm-N(T, t) all p-values required are calculated and can be cached by Algorithm-N(T, t) F and F 1 procedures are linear Correctly control/bound the FDR under the assumptions: A1: the network is t-faithful(t ) A2: the FDR procedures F and F 1 are correct in BN Learning Ioannis Tsamardinos, Univ. of Crete 41 / 78
42 in BN Learning Ioannis Tsamardinos, Univ. of Crete 42 / 78
43 V (d): # of false positives when we reject all hypotheses with a p-values less than d R(d): # of positives (rejections) when we reject all hypotheses with a p-value less than d FDR(d) = Ex(V (d)/r(d)) in BN Learning Ioannis Tsamardinos, Univ. of Crete 43 / 78
44 Approximating FDR(d) Approximation 1: Ex(V (d)/r(d)) Ex(V (d))/e(r(d)) Approximation 2: Ex(V (d)) d m, where m = V \{T } Approximation 3: Ex(R(d)) R o (d), where R o (d) is the observed number of rejections (i.e. N(T, t, d) ) in BN Learning Ioannis Tsamardinos, Univ. of Crete 44 / 78
45 Approximating FDR(d) Ex(V (d)) d m, where m = V \{T } in the worst case all m hypotheses are truly null P(p i,t d null is true) = d So, on average in the worst case we will reject d m truly nulls in BN Learning Ioannis Tsamardinos, Univ. of Crete 45 / 78
46 FDR Approximation FDR(d) =Ex(V (d)/r(d)) a m d/r 0 (d) a m d/#{p i,t d} a F({p i,t }, a) = p k,t where p k,t is the largest p-value such that FDR(p k,t ) a F 1 ({p i,t }, d) = FDR(p k,t ) FDR(d) where p k,t is the largest p-value such that p k,t d in BN Learning Ioannis Tsamardinos, Univ. of Crete 46 / 78
47 C-FDR(T, t,a): Control FDR Algorithm 7 C-FDR(T, t, a) 1: procedure C-FDR(T, t, a) 2: N(T) = Algorithm-N(T,t) 3: for all do V i N(T) 4: E(V i )= Algorithm-E(V i,t) 5: p i,t = max{p i,t,k,z k N(T) or Z k E(V i )} 6: end for 7: 8: d = F({p i,t },a) return {V i : p i,t d} and FDR(d) 9: end procedure FDR(d) a: FDR(d) the largest FDR smaller than a that could be achieved (called Estimated Alpha) in BN Learning Ioannis Tsamardinos, Univ. of Crete 47 / 78
48 in BN Learning Ioannis Tsamardinos, Univ. of Crete 48 / 78
49 Experimental setup Network Total Vars Selected Alarm Child Insurance Pigs Gene variables 5 samplings = 790 runs per sample size and threshold setting in BN Learning Ioannis Tsamardinos, Univ. of Crete 49 / 78
50 Algorithm Evaluated B-FDR(T, t, t) Returns a bound on the FDR for a given target and dataset Bound is provided assuming t-faithfulness(t ) and rejecting a hypothesis if p i,t < t Instantiation of Algorithm-N(T, t) used is the Max-Min Parents and Children in BN Learning Ioannis Tsamardinos, Univ. of Crete 50 / 78
51 Results: Sample Size Effect Empirical FDR on MMPC Output Sample Size 1000, Alpha = (a) Sample Size 5000, Alpha = (b) Child Insurance Alarm Pigs Gene Estimated Bound on FDR Sample Size 10000, Alpha = (c) Average bound over samplings vs average true FDR in BN Learning Ioannis Tsamardinos, Univ. of Crete 51 / 78
52 Results: Threshold Effect t Sample Size 5000, Alpha = Sample Size 5000, Alpha = Empirical FDR on MMPC Output (a) Sample Size 5000, Alpha = (b) Sample Size 5000, Alpha = Child Insurance Alarm Pigs Gene (c) Estimated Bound on FDR (d) in BN Learning Ioannis Tsamardinos, Univ. of Crete 52 / 78
53 Quantitative Results SS Alpha Num UE Avg. Error Avg Slack Error = max(0, true-fdr FDR-bound) Slack = max(0, FDR-bound true-fdr) UE: experiment where bound was underestimated in BN Learning Ioannis Tsamardinos, Univ. of Crete 53 / 78
54 Algorithm Evaluated C-FDR(T, t, a) Instantiation of Algorithm-N(T, t) used is the Max-Min Parents and Children Threshold t set to 0.15 a ranged in { 10%, 20%,..., 90% } We compare the Estimated Alpha returned with the true FDR achieved in BN Learning Ioannis Tsamardinos, Univ. of Crete 54 / 78
55 Results: Alpha Effect in BN Learning Ioannis Tsamardinos, Univ. of Crete 55 / 78
56 Execution Time Average time to run Algorithm-N(T, t), B-FDR(T, t, d), C-FDR(T, t, a) 33 seconds Excluding the Pigs network the average drops to 12 seconds in BN Learning Ioannis Tsamardinos, Univ. of Crete 56 / 78
57 in BN Learning Ioannis Tsamardinos, Univ. of Crete 57 / 78
58 Effect of Small R o (d) FDR(d) =Ex(V (d)/r(d)) a m d/r 0 (d) a m d/#{p i,t d} a When R o (d) is small, a small difference dramatically changes the FDR estimate Increases variance in BN Learning Ioannis Tsamardinos, Univ. of Crete 58 / 78
59 Variance vs R 0 (d) STD(Bound FDR Empirical FDR) Number of Rejections in BN Learning Ioannis Tsamardinos, Univ. of Crete 59 / 78
60 Dealing with small R o (d) Make no claim when R o (d) is small Increase R o (d) somehow Solve a learning problem where R o (d) is relatively large Learn the whole skeleton of the network (all the edges) Similar theory, similar algorithms Or, learn a Region of Interest around T, e.g., the region at most 3 edges away from T in BN Learning Ioannis Tsamardinos, Univ. of Crete 60 / 78
61 C-FDR(t,a): control FDR Algorithm 8 C-FDR(t, a) 1: procedure C-FDR(t, a) 2: for all V i V do 3: E(V i )=Algorithm-N(V i, T) 4: end for 5: for all V i and V j E(V j ) do 6: p i,j = max{p i,j,k,z k E(V i ) or Z k E(V j )} 7: end for 8: 9: d = F(p i,j, i < j, a) % because of symmetry return {edge(v i, V j ) : p i,j d} 10: end procedure The number of tests/hypotheses are n(n 1)/2, i.e., the number of possible edges in BN Learning Ioannis Tsamardinos, Univ. of Crete 61 / 78
62 B-FDR(t, d): Bound FDR Bound the FDR of learning the edges in the network assuming t-faithful(t ) and rejecting a complex hypothesis when p i,j d Algorithm 9 B-FDR(t, a) 1: procedure B-FDR(t, a) 2: for all variable V i do 3: E(V i )=Algorithm-N(V i, T) 4: end for 5: for all V i and V j E(V j ) do 6: p i,j = max{p i,j,k,z k E(V i ) or Z k E(V j )} 7: end for 8: a = F 1 (p i,j, i < j, d) 9: return a 10: end procedure in BN Learning Ioannis Tsamardinos, Univ. of Crete 62 / 78
63 a Test in BN Learning Ioannis Tsamardinos, Univ. of Crete 63 / 78
64 The Reality About Tests of Independence When p i,j,k is large it may be that The independence I(V i ; V j Z k ) holds indeed There is not enough sample to obtain a statistically significant results (low power) The larger the Z k the lower the power If we perform all possible tests, some will return a high p i,j,k because of low power and we will believe we have found a certificate of exclusion in BN Learning Ioannis Tsamardinos, Univ. of Crete 64 / 78
65 The Reality About Tests of Independence All algorithms based on tests of independence only perform them when they consider they believe they have enough statistical power PC/MMPC Only perform a test T(V i ; V j Z k ) when there are at least 10/5 samples per cell of the contingency tables It is possible, an important test T(V i ; V j Z k ) not to be performed in BN Learning Ioannis Tsamardinos, Univ. of Crete 65 / 78
66 Assumptions Assumption 1: the network is t-faithful(t ) Assumption 2: the FDR procedure is correct Assumption 3: All tests required by the algorithms and containing certificates of exclusion tests will be performed in BN Learning Ioannis Tsamardinos, Univ. of Crete 66 / 78
67 Sample Size 1000 Experiment Empirical FDR on MMPC output Sample Size 1000, Alpha Threshold = 0.05 Child Insurance Alarm Pigs Gene Estimated Bound on FDR Empirical FDR on MMPC output Filtered Sample Size 1000 Child Insurance Alarm Pigs Gene Estimated Bound on FDR SS Alpha Num UE Avg. Error Avg. Slack Original Filtered Experiments where Assumption 3 is violated are filtered out in BN Learning Ioannis Tsamardinos, Univ. of Crete 67 / 78
68 Guarantees in Small Sample Size Suppose that there is not enough sample to condition on large enough sets Then, we expect some structural false positives within the returned estimate of N(T) What can we do? Provide looser guarantees Change the definition of False Discovery in BN Learning Ioannis Tsamardinos, Univ. of Crete 68 / 78
69 False Discovery Definition False Discovery: V i is returned in N(T, t, a) but V i / N(T) FD(T, k) set of V i such that V i V, Z N(T) or Z N(V i ) with Z k, I(V i, T Z) When k max( N(T), N(V i) ), then FD(T, k) = N(T) in BN Learning Ioannis Tsamardinos, Univ. of Crete 69 / 78
70 False Discovery Definition Before: Now: False Discovery: V i is returned in N(T,t,a) but V i / N(T) False Discovery: V i is returned in N(T,t,a) but V i / FD(T,k) in BN Learning Ioannis Tsamardinos, Univ. of Crete 70 / 78
71 Using FD(T,k) Assume the maximum conditioning size is 1 We can d-separate V 3 from T conditioned on 2 variables, but not just 1 variable V 3 is a false discovery with our previous definition V 3 FD(T, 1) and so is not a false discovery 1 using the new definition in BN Learning Ioannis Tsamardinos, Univ. of Crete 71 / 78
72 Using FD(T,k) Suppose we determine that with the given dataset we can reliably condition on at most k variables Assuming: the network is t-faithful(t) the FDR procedure is correct We guarantee that FDR of N(T, t, a) a, where a False Discovery is defined as V i / FD(T, k) in BN Learning Ioannis Tsamardinos, Univ. of Crete 72 / 78
73 in BN Learning Ioannis Tsamardinos, Univ. of Crete 73 / 78
74 Ideas For Extensions Approximation 2: Ex(V (d)) d m Ex(V (d)) d m 0 m 0 is an estimate of the true null hypotheses; one could use Storey s procedures to estimate m 0 Estimate Ex(V (d)) by permutation testing Permute the data of T, run the Algorithm-N(T,t) and measure the size of N(T,t) which are all false positives by construction Measure the average size of N(T,t) in all permutations and use it as Ex(V (t)) in BN Learning Ioannis Tsamardinos, Univ. of Crete 74 / 78
75 Ideas For Extensions What about false negatives? there are statistical procedures that could be applied What about other tasks? Orientation of edges based on independence testing Discovery of hidden variables in BN Learning Ioannis Tsamardinos, Univ. of Crete 75 / 78
76 Discussion A step towards a theory of causal discovery and BN learning in the finite sample Algorithms work well when the assumptions hold and sample is relatively large Statistical guarantees produced should be intuitive to practitioners in BN Learning Ioannis Tsamardinos, Univ. of Crete 76 / 78
77 Limitations p-values are only approximated and FDR procedures used are conservative Leads to reduced power of the complex hypotheses (increased false negatives) Further evaluation is needed especially for untested algorithms and extensions over a larger range of networks and parameters with particular attention to the variance of the FDR bounds comparison with other related work [5] in BN Learning Ioannis Tsamardinos, Univ. of Crete 77 / 78
78 References [1] Y. Benjamini and Y. Hochberg. false discovery rate - a practical and powerful approach to multiple testing. J Roy Stat Soc B Methods, 57(1): , [2] D. M. Chickering. Optimal structure identification with greedy search. Journal of Machine Learning Research, pages , [3] Nir Friedman and Daphne Koller. Being bayesian about network structure. pages [4] Markus Kalisch and Peter Bühlmann. Estimating high-dimensional directed acyclic graphs with the pc-algorithm. J. Mach. Learn. Res., 8: , [5] J. Listgarten and D. Heckerman. Determining the number of non-spurious arcs in a learned dag model: Investigation of a bayesian and a frequentist approach. In 23rd Conference on Uncertainty in Artificial Intelligence, [6] P. Spirtes, C. Glymour, and R. Scheines. Causation, Prediction, and Search. MIT Press, 2nd edition, [7] I Tsamardinos, LE Brown, and CF Aliferis. The Max-Min Hill-Climbing bayesian network structure learning algorithm. Machine Learning, 65(1):31 78, [8] Ioannis Tsamardinos and Laura E. Brown. Bounding the false discovery rate in local bayesian in BN Learning Ioannis Tsamardinos, Univ. of Crete 78 / 78 network learning. In AAAI, 2008.
ONLINE DELAY MANAGEMENT IN RAILWAYS - SIMULATION OF A TRAIN TIMETABLE
ONLINE DELAY MANAGEMENT IN RAILWAYS - SIMULATION OF A TRAIN TIMETABLE WITH DECISION RULES - N. VAN MEERTEN 333485 28-08-2013 Econometrics & Operational Research Erasmus University Rotterdam Bachelor thesis
More informationUC Berkeley Working Papers
UC Berkeley Working Papers Title The Value Of Runway Time Slots For Airlines Permalink https://escholarship.org/uc/item/69t9v6qb Authors Cao, Jia-ming Kanafani, Adib Publication Date 1997-05-01 escholarship.org
More informationPredicting a Dramatic Contraction in the 10-Year Passenger Demand
Predicting a Dramatic Contraction in the 10-Year Passenger Demand Daniel Y. Suh Megan S. Ryerson University of Pennsylvania 6/29/2018 8 th International Conference on Research in Air Transportation Outline
More informationAbstract. Introduction
COMPARISON OF EFFICIENCY OF SLOT ALLOCATION BY CONGESTION PRICING AND RATION BY SCHEDULE Saba Neyshaboury,Vivek Kumar, Lance Sherry, Karla Hoffman Center for Air Transportation Systems Research (CATSR)
More informationCross-sectional time-series analysis of airspace capacity in Europe
Cross-sectional time-series analysis of airspace capacity in Europe Dr. A. Majumdar Dr. W.Y. Ochieng Gerard McAuley (EUROCONTROL) Jean Michel Lenzi (EUROCONTROL) Catalin Lepadatu (EUROCONTROL) 1 Introduction
More informationA Statistical Method for Eliminating False Counts Due to Debris, Using Automated Visual Inspection for Probe Marks
A Statistical Method for Eliminating False Counts Due to Debris, Using Automated Visual Inspection for Probe Marks SWTW 2003 Max Guest & Mike Clay August Technology, Plano, TX Probe Debris & Challenges
More informationAn Analytical Approach to the BFS vs. DFS Algorithm Selection Problem 1
An Analytical Approach to the BFS vs. DFS Algorithm Selection Problem 1 Tom Everitt Marcus Hutter Australian National University September 3, 2015 Everitt, T. and Hutter, M. (2015a). Analytical Results
More informationB.S. PROGRAM IN AVIATION TECHNOLOGY MANAGEMENT Course Descriptions
Course Descriptions 01225111 Basic Mathematics in Aviation 3(3-0-6) Algebra. Functions and graphs. Limit and continuity. Derivatives. Integration. Applications in aviation technology management. 01225121
More informationAirline Scheduling Optimization ( Chapter 7 I)
Airline Scheduling Optimization ( Chapter 7 I) Vivek Kumar (Research Associate, CATSR/GMU) February 28 th, 2011 CENTER FOR AIR TRANSPORTATION SYSTEMS RESEARCH 2 Agenda Airline Scheduling Factors affecting
More informationA Duality Based Approach for Network Revenue Management in Airline Alliances
A Duality Based Approach for Network Revenue Management in Airline Alliances Huseyin Topaloglu School of Operations Research and Information Engineering, Cornell University, Ithaca, New York 14853, USA
More informationSimulation of disturbances and modelling of expected train passenger delays
Computers in Railways X 521 Simulation of disturbances and modelling of expected train passenger delays A. Landex & O. A. Nielsen Centre for Traffic and Transport, Technical University of Denmark, Denmark
More informationAmerican Airlines Next Top Model
Page 1 of 12 American Airlines Next Top Model Introduction Airlines employ several distinct strategies for the boarding and deboarding of airplanes in an attempt to minimize the time each plane spends
More informationCHAPTER 5 SIMULATION MODEL TO DETERMINE FREQUENCY OF A SINGLE BUS ROUTE WITH SINGLE AND MULTIPLE HEADWAYS
91 CHAPTER 5 SIMULATION MODEL TO DETERMINE FREQUENCY OF A SINGLE BUS ROUTE WITH SINGLE AND MULTIPLE HEADWAYS 5.1 INTRODUCTION In chapter 4, from the evaluation of routes and the sensitive analysis, it
More informationTowards New Metrics Assessing Air Traffic Network Interactions
Towards New Metrics Assessing Air Traffic Network Interactions Silvia Zaoli Salzburg 6 of December 2018 Domino Project Aim: assessing the impact of innovations in the European ATM system Innovations change
More informationIncluding Linear Holding in Air Traffic Flow Management for Flexible Delay Handling
Including Linear Holding in Air Traffic Flow Management for Flexible Delay Handling Yan Xu and Xavier Prats Technical University of Catalonia (UPC) Outline Motivation & Background Trajectory optimization
More informationHOW TO IMPROVE HIGH-FREQUENCY BUS SERVICE RELIABILITY THROUGH SCHEDULING
HOW TO IMPROVE HIGH-FREQUENCY BUS SERVICE RELIABILITY THROUGH SCHEDULING Ms. Grace Fattouche Abstract This paper outlines a scheduling process for improving high-frequency bus service reliability based
More informationA RECURSION EVENT-DRIVEN MODEL TO SOLVE THE SINGLE AIRPORT GROUND-HOLDING PROBLEM
RECURSION EVENT-DRIVEN MODEL TO SOLVE THE SINGLE IRPORT GROUND-HOLDING PROBLEM Lili WNG Doctor ir Traffic Management College Civil viation University of China 00 Xunhai Road, Dongli District, Tianjin P.R.
More informationCOMPARATIVE STUDY ON GROWTH AND FINANCIAL PERFORMANCE OF JET AIRWAYS, INDIGO AIRLINES & SPICEJET AIRLINES COMPANIES IN INDIA
Volume 2, Issue 2, November 2017, ISBR Management Journal ISSN(Online)- 2456-9062 COMPARATIVE STUDY ON GROWTH AND FINANCIAL PERFORMANCE OF JET AIRWAYS, INDIGO AIRLINES & SPICEJET AIRLINES COMPANIES IN
More informationADVANTAGES OF SIMULATION
ADVANTAGES OF SIMULATION Most complex, real-world systems with stochastic elements cannot be accurately described by a mathematical model that can be evaluated analytically. Thus, a simulation is often
More informationIdentification of Waves in IGC files
Identification of Waves in IGC files Prof. Dr. Databionics Research Group University of Marburg Databionics Databionics means copying algorithms from nature e.g. swarm algorithms, neural networks, social
More informationGenetic Algorithm in Python. Data mining lab 6
Genetic Algorithm in Python Data mining lab 6 When to use genetic algorithms John Holland (1975) Optimization: minimize (maximize) some function f(x) over all possible values of variables x in X A brute
More informationQuantile Regression Based Estimation of Statistical Contingency Fuel. Lei Kang, Mark Hansen June 29, 2017
Quantile Regression Based Estimation of Statistical Contingency Fuel Lei Kang, Mark Hansen June 29, 2017 Agenda Background Industry practice Data Methodology Benefit assessment Conclusion 2 Agenda Background
More informationThe Effectiveness of JetBlue if Allowed to Manage More of its Resources
McNair Scholars Research Journal Volume 2 Article 4 2015 The Effectiveness of JetBlue if Allowed to Manage More of its Resources Jerre F. Johnson Embry Riddle Aeronautical University, johnsff9@my.erau.edu
More informationATM Seminar 2015 OPTIMIZING INTEGRATED ARRIVAL, DEPARTURE AND SURFACE OPERATIONS UNDER UNCERTAINTY. Wednesday, June 24 nd 2015
OPTIMIZING INTEGRATED ARRIVAL, DEPARTURE AND SURFACE OPERATIONS UNDER UNCERTAINTY Christabelle Bosson PhD Candidate Purdue AAE Min Xue University Affiliated Research Center Shannon Zelinski NASA Ames Research
More informationAQME 10 System Description
AQME 10 System Description Luca Pulina and Armando Tacchella University of Genoa DIST - Viale Causa 13 16145 Genoa (Italy) POS 2010 - Edinburgh, July 10, 2010 Luca Pulina (UNIGE) AQME 10 System Description
More informationThree Essays on the Introduction and Impact of Baggage Fees in the U.S. Airline Industry
Clemson University TigerPrints All Dissertations Dissertations 5-2016 Three Essays on the Introduction and Impact of Baggage Fees in the U.S. Airline Industry Alexander Fiore Clemson University, afiore@g.clemson.edu
More informationAIRLINES decisions on route selection are, along with fleet planning and schedule development, the most important
Modeling Airline Decisions on Route Planning Using Discrete Choice Models Zhenghui Sha, Kushal Moolchandani, Apoorv Maheshwari, Joseph Thekinen, Jitesh H. Panchal, Daniel A. DeLaurentis Purdue University,
More informationDecision aid methodologies in transportation
Decision aid methodologies in transportation Lecture 5: Revenue Management Prem Kumar prem.viswanathan@epfl.ch Transport and Mobility Laboratory * Presentation materials in this course uses some slides
More informationHeuristic technique for tour package models
Proceedings of the 214 International Conference on Information, Operations Management and Statistics (ICIOMS213), Kuala Lumpur, Malaysia, September 1-3, 213 Heuristic technique for tour package models
More informationA. Karakasidis 1, V. S. Verykios 2 and P. Christen 3
A. Karakasidis 1, V. S. Verykios 2 and P. Christen 3 1 Department of Computer and Communication Engineering University of Thessaly Volos, Greece akarakasidis@inf.uth.gr 2 School of Science and Technology
More informationDo Not Write Below Question Maximum Possible Points Score Total Points = 100
University of Toronto Department of Economics ECO 204 Summer 2012 Ajaz Hussain TEST 3 SOLUTIONS TIME: 1 HOUR AND 50 MINUTES YOU CANNOT LEAVE THE EXAM ROOM DURING THE LAST 10 MINUTES OF THE TEST. PLEASE
More informationCongestion. Vikrant Vaze Prof. Cynthia Barnhart. Department of Civil and Environmental Engineering Massachusetts Institute of Technology
Frequency Competition and Congestion Vikrant Vaze Prof. Cynthia Barnhart Department of Civil and Environmental Engineering Massachusetts Institute of Technology Delays and Demand Capacity Imbalance Estimated
More informationAn Examination of the Effect of Multiple Supervisors on Flight Trainees' Performance
National Training Aircraft Symposium (NTAS) 2018 - The Changing Role of the Pilot Aug 14th, 10:30 AM - 11:45 AM An Examination of the Effect of Multiple Supervisors on Flight Trainees' Performance Dongyun
More informationComparative Densities of Tigers (Panthera tigris tigris) between Tourism and Non Tourism Zone of Pench Tiger Reserve, Madhya Pradesh- A brief report
Comparative Densities of Tigers (Panthera tigris tigris) between Tourism and Non Tourism Zone of Pench Tiger Reserve, Madhya Pradesh- A brief report Submitted by Principal investigators Prof. (Dr.) K.
More information= Coordination with Direct Communication
Particle Swarm Optimization Mohamed A. El-Sharkawi Computational Intelligence Applications (CIA) Lab. Department of EE, Box 352500 University of Washington Seattle, WA 98195-2500 elsharkawi@ee.washington.edu
More informationWhite Paper: Assessment of 1-to-Many matching in the airport departure process
White Paper: Assessment of 1-to-Many matching in the airport departure process November 2015 rockwellcollins.com Background The airline industry is experiencing significant growth. With higher capacity
More informationTransportation Timetabling
Outline DM87 SCHEDULING, TIMETABLING AND ROUTING Lecture 16 Transportation Timetabling 1. Transportation Timetabling Tanker Scheduling Air Transport Train Timetabling Marco Chiarandini DM87 Scheduling,
More informationBest schedule to utilize the Big Long River
page 1of20 1 Introduction Best schedule to utilize the Big Long River People enjoy going to the Big Long River for its scenic views and exciting white water rapids, and the only way to achieve this should
More informationSchedule Compression by Fair Allocation Methods
Schedule Compression by Fair Allocation Methods by Michael Ball Andrew Churchill David Lovell University of Maryland and NEXTOR, the National Center of Excellence for Aviation Operations Research November
More informationAn Econometric Study of Flight Delay Causes at O Hare International Airport Nathan Daniel Boettcher, Dr. Don Thompson*
An Econometric Study of Flight Delay Causes at O Hare International Airport Nathan Daniel Boettcher, Dr. Don Thompson* Abstract This study examined the relationship between sources of delay and the level
More information15:00 minutes of the scheduled arrival time. As a leader in aviation and air travel data insights, we are uniquely positioned to provide an
FlightGlobal, incorporating FlightStats, On-time Performance Service Awards: A Long-time Partner Recognizing Industry Success ON-TIME PERFORMANCE 2018 WINNER SERVICE AWARDS As a leader in aviation and
More informationResearch on Pilots Development Planning
Journal of Software Engineering and Applications 2012 5 1016-1022 http://dx.doi.org/10.4236/sea.2012.512118 Published Online December 2012 (http://www.scirp.org/ournal/sea) Ruo Ding Mingang Gao * Institute
More informationPERFORMANCE MEASURE INFORMATION SHEET #16
PERFORMANCE MEASURE INFORMATION SHEET #16 ARROW LAKES RESERVOIR: RECREATION Objective / Location Recreation/Arrow Lakes Reservoir Performance Measure Access Days Units Description MSIC 1) # Access Days
More informationTwo Major Problems Problems Crew Pairing Problem (CPP) Find a set of legal pairin Find gs (each pairing
Solving Airline s Pilot-Copilot Rostering Problem by Successive Bipartite Weighted Matching by Xugang Ye Applied Mathematics and Statistics, The Johns Hopkins University Motivation Crew-related related
More informationSERVICE NETWORK DESIGN: APPLICATIONS IN TRANSPORTATION AND LOGISTICS
SERVICE NETWORK DESIGN: APPLICATIONS IN TRANSPORTATION AND LOGISTICS Professor Cynthia Barnhart Massachusetts Institute of Technology Cambridge, Massachusetts USA March 21, 2007 Outline Service network
More informationDon t Sit on the Fence
Don t Sit on the Fence A Static Analysis Approach to Automatic Fence Insertion Or Ostrovsky April 25th 2018 Or Ostrovsky Don t Sit on the Fence April 25th 2018 1 / 50 Table of contents 1 Introduction 2
More informationATTEND Analytical Tools To Evaluate Negotiation Difficulty
ATTEND Analytical Tools To Evaluate Negotiation Difficulty Alejandro Bugacov Robert Neches University of Southern California Information Sciences Institute ANTs PI Meeting, November, 2000 Outline 1. Goals
More informationImpact of Landing Fee Policy on Airlines Service Decisions, Financial Performance and Airport Congestion
Wenbin Wei Impact of Landing Fee Policy on Airlines Service Decisions, Financial Performance and Airport Congestion Wenbin Wei Department of Aviation and Technology San Jose State University One Washington
More informationAirspace Complexity Measurement: An Air Traffic Control Simulation Analysis
Airspace Complexity Measurement: An Air Traffic Control Simulation Analysis Parimal Kopardekar NASA Ames Research Center Albert Schwartz, Sherri Magyarits, and Jessica Rhodes FAA William J. Hughes Technical
More informationPredicting flight routes with a Deep Neural Network in the operational Air Traffic Flow and Capacity Management system
FEB 2018 EUROCONTROL Maastricht Upper Area Control Centre Predicting flight routes with a Deep Neural Network in the operational Air Traffic Flow and Capacity Management system Trajectory prediction is
More informationThanksgiving Holiday Period Traffic Fatality Estimate, 2017
Thanksgiving Holiday Period Traffic Fatality Estimate, 2017 Prepared by Research and Statistics Department National Safety Council November 13, 2017 Holiday period definition Thanksgiving is observed on
More informationArash Yousefi George L. Donohue, Ph.D. Chun-Hung Chen, Ph.D.
Investigation of Airspace Metrics for Design and Evaluation of New ATM Concepts Arash Yousefi George L. Donohue, Ph.D. Chun-Hung Chen, Ph.D. Air Transportation Systems Lab George Mason University Presented
More informationForegone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035
Foregone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035 Foregone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035 George Anjaparidze IATA, February 2015 Version1.1
More informationPerformance Indicator Horizontal Flight Efficiency
Performance Indicator Horizontal Flight Efficiency Level 1 and 2 documentation of the Horizontal Flight Efficiency key performance indicators Overview This document is a template for a Level 1 & Level
More informationTactical and Operational Planning of Scheduled Maintenance for Per-Seat, On-Demand Air Transportation
Tactical and Operational Planning of Scheduled Maintenance for Per-Seat, On-Demand Air Transportation Gizem Keysan, George L. Nemhauser, and Martin W.P. Savelsbergh February 13, 2009 Abstract Advances
More informationEvaluation of Predictability as a Performance Measure
Evaluation of Predictability as a Performance Measure Presented by: Mark Hansen, UC Berkeley Global Challenges Workshop February 12, 2015 With Assistance From: John Gulding, FAA Lu Hao, Lei Kang, Yi Liu,
More informationPerth & Kinross Council. Community Planning Partnership Report June 2016
Perth & Kinross Council Community Planning Partnership Report June 2016 Contents Foreword... 3 Section 1: Spring 2016 destination follow up of 2014/15 school leavers... 4 Background... 4 Section A: Initial
More informationLogic Control Summer Semester Assignment: Modeling and Logic Controller Design 1
TECHNISCHE UNIVERSITÄT DORTMUND Faculty of Bio- and Chemical Engineering Process Dynamics and Operations Group Prof. Dr.-Ing. Sebastian Engell D Y N Logic Control Summer Semester 2018 Assignment: Modeling
More informationSemantic Representation and Scale-up of Integrated Air Traffic Management Data
Semantic Representation and Scale-up of Integrated Air Traffic Management Data Rich Keller, Ph.D. * Mei Wei * Shubha Ranjan + Michelle Eshow *Intelligent Systems Division / Aviation Systems Division +
More informationAssignment of Arrival Slots
Assignment of Arrival Slots James Schummer Rakesh V. Vohra Kellogg School of Management (MEDS) Northwestern University March 2012 Schummer & Vohra (Northwestern Univ.) Assignment of Arrival Slots March
More informationUsing PBN for Terminal and Extended Terminal Operations
Using PBN for Terminal and Extended Terminal Operations Navigation Performance Data Analysis and its Effect on Route Spacing Dijana Trenevska EUROCONTROL 27 June 2017 Content Background and Objective Data
More informationQuantitative Analysis of the Adapted Physical Education Employment Market in Higher Education
Quantitative Analysis of the Adapted Physical Education Employment Market in Higher Education by Jiabei Zhang, Western Michigan University Abstract The purpose of this study was to analyze the employment
More informationResearch Article Study on Fleet Assignment Problem Model and Algorithm
Mathematical Problems in Engineering Volume 2013, Article ID 581586, 5 pages http://dxdoiorg/101155/2013/581586 Research Article Study on Fleet Assignment Problem Model and Algorithm Yaohua Li and Na Tan
More informationCONNECT Events: Flight Optimization
CONNECT Events: Flight Optimization Ian Britchford Director Post Flight Solutions 5 th October 2016 Data Analysis and Root Cause Evaluation for Continuous Improvement Learn about Jeppesen s next level
More informationBioinformatics of Protein Domains: New Computational Approach for the Detection of Protein Domains
Bioinformatics of Protein Domains: New Computational Approach for the Detection of Protein Domains Maricel Kann Assistant Professor University of Maryland, Baltimore County mkann@umbc.edu Maricel Kann.
More informationActivity Template. Drexel-SDP GK-12 ACTIVITY. Subject Area(s): Sound Associated Unit: Associated Lesson: None
Activity Template Subject Area(s): Sound Associated Unit: Associated Lesson: None Drexel-SDP GK-12 ACTIVITY Activity Title: What is the quickest way to my destination? Grade Level: 8 (7-9) Activity Dependency:
More informationResponse of U.S. Air Carriers to On-Time Disclosure Rule
TRANSPORTATION RESEARCH RECORD 1379 9 Response of US Air Carriers to On-Time Disclosure Rule ROBERT SHUMSKY The On-Time Disclosure Rule, implemented by the US Department of Transportation in 1987, makes
More informationCluster A.2: Linear Functions, Equations, and Inequalities
A.2A: Representing Domain and Range Values: Taxi Trips Focusing TEKS A.2A Linear Functions, Equations, and Inequalities. The student applies mathematical process standards when using properties of linear
More informationEstimating the Risk of a New Launch Vehicle Using Historical Design Element Data
International Journal of Performability Engineering, Vol. 9, No. 6, November 2013, pp. 599-608. RAMS Consultants Printed in India Estimating the Risk of a New Launch Vehicle Using Historical Design Element
More informationOPTIMAL PUSHBACK TIME WITH EXISTING UNCERTAINTIES AT BUSY AIRPORT
OPTIMAL PUSHBACK TIME WITH EXISTING Ryota Mori* *Electronic Navigation Research Institute Keywords: TSAT, reinforcement learning, uncertainty Abstract Pushback time management of departure aircraft is
More informationOptimal assignment of incoming flights to baggage carousels at airports
Downloaded from orbit.dtu.dk on: May 05, 2018 Optimal assignment of incoming flights to baggage carousels at airports Barth, Torben C. Publication date: 2013 Document Version Publisher's PDF, also known
More informationOperational Performance of High-Occupancy Vehicle (HOV) Facilities:
0 0 Operational Performance of High-Occupancy Vehicle (HOV) Facilities: Comparison of Contiguous and Limited Access HOV Lanes in California Kitae Jang, Ph.D. The Cho Chun Shik Graduate School for Green
More informationChapter 9 Validation Experiments
Chapter 9 Validation Experiments The variable rate model developed for MH37 was validated by analysing data from a collection of flights where the true aircraft location was known; we refer to these as
More informationRUNWAY OPERATIONS: Computing Runway Arrival Capacity
RUNWAY OPERATIONS: Computing Runway Arrival Capacity SYST 560/460 USE Runway Capacity Spreadsheet Fall 2008 Lance Sherry 1 CENTER FOR AIR TRANSPORTATION SYSTEMS RESEARCH Background Air Transportation System
More informationReducing Garbage-In for Discrete Choice Model Estimation
Reducing Garbage-In for Discrete Choice Model Estimation David Kurth* Cambridge Systematics, Inc. 999 18th Street, Suite 3000 Denver, CO 80202 P: 303-357-4661 F: 303-446-9111 dkurth@camsys.com Marty Milkovits
More informationDemand, Load and Spill Analysis Dr. Peter Belobaba
Demand, Load and Spill Analysis Dr. Peter Belobaba Istanbul Technical University Air Transportation Management M.Sc. Program Network, Fleet and Schedule Strategic Planning Module 13 : 12 March 2014 Lecture
More informationPHY 133 Lab 6 - Conservation of Momentum
Stony Brook Physics Laboratory Manuals PHY 133 Lab 6 - Conservation of Momentum The purpose of this lab is to demonstrate conservation of linear momentum in one-dimensional collisions of objects, and to
More informationMIT ICAT M I T I n t e r n a t i o n a l C e n t e r f o r A i r T r a n s p o r t a t i o n
M I T I n t e r n a t i o n a l C e n t e r f o r A i r T r a n s p o r t a t i o n PRICING AND REVENUE MANAGEMENT RESEARCH Airline Competition and Pricing Power Presentations to Industry Advisory Board
More informationResearch on Aviation Security*
Research on Aviation Security* John E. Kobza Department of Industrial Engineering Texas Tech University (806) 742-3402 john.kobza@coe.ttu.edu www.coe.ttu.edu/ie/kobza.htm Sheldon H. Jacobson Department
More informationA comparison of two methods for reducing take-off delay at London Heathrow airport
MISTA 2009 A comparison of two methods for reducing take-off delay at London Heathrow airport Jason A. D. Atkin Edmund K. Burke John S Greenwood Abstract This paper describes recent research into the departure
More informationTicket reservation posts on train platforms: an assessment using the microscopic pedestrian simulation tool Nomad
Daamen, Hoogendoorn, Campanella and Eggengoor 1 Ticket reservation posts on train platforms: an assessment using the microscopic pedestrian simulation tool Nomad Winnie Daamen, PhD (corresponding author)
More informationIPSOS / REUTERS POLL DATA Prepared by Ipsos Public Affairs
Ipsos Poll Conducted for Reuters Airlines Poll 6.30.2017 These are findings from an Ipsos poll conducted June 22-29, 2017 on behalf Thomson Reuters. For the survey, a sample of roughly 2,316 adults age
More informationMotion 2. 1 Purpose. 2 Theory
Motion 2 Equipment Capstone, motion sensor, meter stick, air track+ 2 gliders, 2 blocks, and index cards. Air Tracks In this experiment you will be using an air track. This is a long straight triangular
More informationESD Working Paper Series
ESD Working Paper Series Airport Congestion Mitigation through Dynamic Control of Runway Configurations and of Arrival and Departure Service Rates under Stochastic Operating Conditions Alexandre Jacquillat
More informationRunway Length Analysis Prescott Municipal Airport
APPENDIX 2 Runway Length Analysis Prescott Municipal Airport May 11, 2009 Version 2 (draft) Table of Contents Introduction... 1-1 Section 1 Purpose & Need... 1-2 Section 2 Design Standards...1-3 Section
More informationChangiNOW: A mobile application for efficient taxi allocation at airports
ChangiNOW: A mobile application for efficient taxi allocation at airports The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation
More informationRECEDING HORIZON CONTROL FOR AIRPORT CAPACITY MANAGEMENT
RECEDING HORIZON CONTROL FOR AIRPORT CAPACITY MANAGEMENT W.-H. Chen, X.B. Hu Dept. of Aeronautical & Automotive Engineering, Loughborough University, UK Keywords: Receding Horizon Control, Air Traffic
More informationSemi - Annual Report. April 2, From September 21, 2003 to March 20, 2004
Comparison of the Effectiveness of a Personal Computer Aviation Training Device, a Flight Training Device, and an Airplane in Conducting Instrument Proficiency Checks Semi - Annual Report April 2, 2004
More information(2, 3) 2 B1 cao. 1 B1 cao
1. 2 1 3 3,,, 5 2 5 4 2. (a) (i) (2, 3) 2 B1 cao (ii) ( 3, 1) B1 cao 3 M1 for correct method to change two fractions to marks or percentages or fractions with a common denominator or decimals with at least
More informationAnalysis and Evaluation of the Slugging Form of Ridesharing*
Analysis and Evaluation of the Slugging Form of Ridesharing* Shuo Ma Department of Compute Science University of Illinois at Chicago Chicago, U.S.A. sma21@uic.edu ABSTRACT Ridesharing is a promising method
More informationModelling Transportation Networks with Octave
Modelling Transportation Networks with Octave Six Silberman June 12, 2008 Abstract This document presents and discusses some code for modelling transportation networks using the Octave language/environment.
More informationPrice-Setting Auctions for Airport Slot Allocation: a Multi-Airport Case Study
Price-Setting Auctions for Airport Slot Allocation: a Multi-Airport Case Study An Agent-Based Computational Economics Approach to Strategic Slot Allocation SESAR Innovation Days Bologna, 2 nd December
More informationIMPROVING THE ROBUSTNESS OF FLIGHT SCHEDULE BY FLIGHT RE-TIMING AND IMPOSING A NEW CREW BASE
Jurnal Karya Asli Lorekan Ahli Matematik Vol. 6 No.1 (2013) Page 066-073. Jurnal Karya Asli Lorekan Ahli Matematik IMPROVING THE ROBUSTNESS OF FLIGHT SCHEDULE BY FLIGHT RE-TIMING AND IMPOSING A NEW CREW
More informationEstimating Sources of Temporal Deviations from Flight Plans
Estimating Sources of Temporal Deviations from Flight Plans Ms. Natasha Yakovchuk (yakovn2@rpi.edu) Prof. Thomas R. Willemain (willet@rpi.edu) Department of Decision Sciences and Engineering Systems Rensselaer
More informationUnit 4: Location-Scale-Based Parametric Distributions
Unit 4: Location-Scale-Based Parametric Distributions Ramón V. León Notes largely based on Statistical Methods for Reliability Data by W.Q. Meeker and L. A. Escobar, Wiley, 1998 and on their class notes.
More informationCAPAN Methodology Sector Capacity Assessment
CAPAN Methodology Sector Capacity Assessment Air Traffic Services System Capacity Seminar/Workshop Nairobi, Kenya, 8 10 June 2016 Raffaele Russo EUROCONTROL Operations Planning Background Network Operations
More informationConstruction of Conflict Free Routes for Aircraft in Case of Free Routing with Genetic Algorithms.
Construction of Conflict Free Routes for Aircraft in Case of Free Routing with Genetic Algorithms. Ingrid Gerdes, German Aerospace Research Establishment, Institute for Flight Guidance, Lilienthalplatz
More informationThe Effects of GPS and Moving Map Displays on Pilot Navigational Awareness While Flying Under VFR
Wright State University CORE Scholar International Symposium on Aviation Psychology - 7 International Symposium on Aviation Psychology 7 The Effects of GPS and Moving Map Displays on Pilot Navigational
More informationTraffic Forecasts. CHAOUKI MUSTAPHA, Economist, International Civil Aviation Organization
SIP/2009-WP/11 Business case Traffic Forecasts CHAOUKI MUSTAPHA, Economist, International Civil Aviation Organization Workshop on the Development of Business Case for the Implementation of CNS/ATM Systems
More informationMAT 115: Precalculus Mathematics Homework Exercises Textbook: A Graphical Approach to Precalculus with Limits: A Unit Circle Approach, Sixth Edition
MAT 115: Precalculus Mathematics Homework Exercises Textbook: A Graphical Approach to Precalculus with Limits: A Unit Circle Approach, Sixth Edition Section R.1, Page 923: Review of Exponents and Polynomials
More information