Estimating the Loss of Efficiency due to Competition in Mobility on Demand Markets
|
|
- Shanon Chapman
- 5 years ago
- Views:
Transcription
1 Estimating the Loss of Efficiency due to Competition in Mobility on Demand Markets Thibault Séjourné 1 Samitha Samaranayake 2 Siddhartha Banerjee 2 1 Ecole Polytechnique 2 Cornell University November 14, 2017 PGMO 2017 November 14, / 38
2 Overview 1 Introduction 2 Worst case study 3 Asymptotic study Description Warm-up: The case of the 2 Nodes Network Results for general networks and consequences PGMO 2017 November 14, / 38
3 Context PGMO 2017 November 14, / 38
4 Context Uber and Lyft drastically reduce the cost of taxis. PGMO 2017 November 14, / 38
5 Context Uber and Lyft drastically reduce the cost of taxis. Other MoD systems are appearing (Velib, Bixi, Car2Go, Autolib). PGMO 2017 November 14, / 38
6 Context Uber and Lyft drastically reduce the cost of taxis. Other MoD systems are appearing (Velib, Bixi, Car2Go, Autolib). 2/3 of the world population living in urban area by PGMO 2017 November 14, / 38
7 Context Uber and Lyft drastically reduce the cost of taxis. Other MoD systems are appearing (Velib, Bixi, Car2Go, Autolib). 2/3 of the world population living in urban area by In the USA, 6.9 bn hours lost in traffic jams, worth $160 bn. PGMO 2017 November 14, / 38
8 Context Uber and Lyft drastically reduce the cost of taxis. Other MoD systems are appearing (Velib, Bixi, Car2Go, Autolib). 2/3 of the world population living in urban area by In the USA, 6.9 bn hours lost in traffic jams, worth $160 bn. MoD systems has a great potential of being an efficient mean of transportation at a cheap price. PGMO 2017 November 14, / 38
9 Issue of competition Let s imagine the case of pooled demands : drop client 2 drop client 1 pick client 1 pick client 2 PGMO 2017 November 14, / 38
10 Issue of competition Let s imagine the case of pooled demands : drop client 2 drop client 1 pick client 1 pick client 2 Competition might considerably undermine the efficiency of MoD systems. PGMO 2017 November 14, / 38
11 Model The city is represented as a complete network of N stations. Hypothesis of an hourly steady state. Distance matrix D = (d ij ) (i,j) and travel time (τ ij ) (i,j) For each edge (i,j), customer demand Λ = (Λ ij ) Node total demand throughput Λ i = j (Λ ji Λ ij ) = (A.Λ) i We study the cost of rebalancing as a function RC(Λ) : min τ ij x ij (i,j) s.t. i, j (x ji x ij ) = Λ i x 0 PGMO 2017 November 14, / 38
12 First properties Properties RC is equal to its dual : RC is convex. max α i.λ i i s.t. (i, j), α i α j τ ij β corner point of the dual, c R, β + c is also a corner point and it yields the same score as β. Consequence: Without loss of generality, we can set any coordinate of β to zero. PGMO 2017 November 14, / 38
13 Worst case study Objective: Find the value of the worst split of demand in two parts, in order to check if the worst case is significant. The Price of Fragmentation (PoF) is: max RC(λ) + RC(Λ λ) s.t. 0 λ Λ PGMO 2017 November 14, / 38
14 Worst case study Objective: Find the value of the worst split of demand in two parts, in order to check if the worst case is significant. The Price of Fragmentation (PoF) is: max RC(λ) + RC(Λ λ) s.t. 0 λ Λ Property The PoF is convex. Thus, the optimal value is reached on a corner point. Proof: Consequence of the convexity of RC. PGMO 2017 November 14, / 38
15 Computing the worst PoF The PoF is a non-linear convex function, almost everywhere differentiable, over a bounded polyhedron isomorph to [0,1]. PGMO 2017 November 14, / 38
16 Computing the worst PoF The PoF is a non-linear convex function, almost everywhere differentiable, over a bounded polyhedron isomorph to [0,1]. Finding the best corner point is NP-hard. PGMO 2017 November 14, / 38
17 Computing the worst PoF The PoF is a non-linear convex function, almost everywhere differentiable, over a bounded polyhedron isomorph to [0,1]. Finding the best corner point is NP-hard. Approximated evaluation via projected subgradient descent. Result: Increase of 567% of the rebalancing cost due to the split. PGMO 2017 November 14, / 38
18 PoF under Stochastic Demand Splitting In practice the worst case will never happen, and there are random fluctuations. PGMO 2017 November 14, / 38
19 PoF under Stochastic Demand Splitting In practice the worst case will never happen, and there are random fluctuations. Use a stochastic model to compute an average loss with respect to the monopoly, ie: γ = E[RC(λ) + RC(Λ λ)] RC(Λ) With λ being a splitting r.v. respecting an exogenous market shares ratio ρ, first assumed to be homogeneous. PGMO 2017 November 14, / 38
20 PoF under Stochastic Demand Splitting In practice the worst case will never happen, and there are random fluctuations. Use a stochastic model to compute an average loss with respect to the monopoly, ie: γ = E[RC(λ) + RC(Λ λ)] RC(Λ) With λ being a splitting r.v. respecting an exogenous market shares ratio ρ, first assumed to be homogeneous. Initial idea: Use a binomial split s.t. λ B (Λ, ρ), assuming Λ has integer coordinates. PGMO 2017 November 14, / 38
21 Properties of the Binomial process Property: Fluid limit Under binomial splitting, Let s define θ N such that λ θ B (θλ, ρ). We have when θ : E[RC(λ θ )]/θ = E[RC(λ θ /θ)] RC(E[λ]) Thus, when the demand is scaled we get that γ 0. Proof: Consequence of the central limit theorem, the continuity of g and the decomposition of binomial into a sum of bernoulli random variables. PGMO 2017 November 14, / 38
22 Consequence of this property The previous property implies that there is no loss of efficiency when the number of demands tends to infinity. Questions If the efficiency loss disappears when the demand is scaled to infinity, then how much does the loss depends on the demand volume? How fast do we converge to the fluid limit? PGMO 2017 November 14, / 38
23 Consequence of this property The previous property implies that there is no loss of efficiency when the number of demands tends to infinity. Questions If the efficiency loss disappears when the demand is scaled to infinity, then how much does the loss depends on the demand volume? How fast do we converge to the fluid limit? We will then focus on the rescaled PoF which depends on θ N: γ θ = PoF (θ) θ = E[RC(λθ ) + RC(θΛ λ θ )] RC(θΛ) θ PGMO 2017 November 14, / 38
24 The example of the 2 nodes network Dual variables for each station α 1 = 0 and α 2, Total demand λ 12 = λ and λ 21 = µ, Random split X N (θρλ, θλs) and Y N (θρµ, θµs), s = ρ(1 ρ) Distances by d 12 = 1 and d 21 = a 1. d λ 12 = λ 12 = λ 21 = µ d 21 = a PGMO 2017 November 14, / 38
25 The example of the 2 nodes network Dual variables for each station α 1 = 0 and α 2, Total demand λ 12 = λ and λ 21 = µ, Random split X N (θρλ, θλs) and Y N (θρµ, θµs), s = ρ(1 ρ) Distances by d 12 = 1 and d 21 = a 1. d λ 12 = λ 12 = λ 21 = µ d 21 = a Formula: RC for one company RC(λ, µ) = max{a.(µ λ), (λ µ)} = [ (a 1).(µ λ) + (a + 1). µ λ ]/2 PGMO 2017 November 14, / 38
26 Two numerical examples Let s set λ = 50, µ = 70, a = 10, ρ = 0.5 Then E[X ] = 25, E[Y ] = 35 γ = g(x, Y ) + g(λ X, µ Y ) g(λ, µ) d λ = = 1 µ = d 21 = 10 PGMO 2017 November 14, / 38
27 Two numerical examples Let s set λ = 50, µ = 70, a = 10, ρ = 0.5 Then E[X ] = 25, E[Y ] = 35 γ = g(x, Y ) + g(λ X, µ Y ) g(λ, µ) d λ = = 1 µ = d 21 = 10 Example 1: X = 30 and Y = 40 Company 1: g(30, 40) = = 100 Company 2: g(20, 30) = = 100 Monopoly: g(50, 70) = 200 PGMO 2017 November 14, / 38
28 Two numerical examples Let s set λ = 50, µ = 70, a = 10, ρ = 0.5 Then E[X ] = 25, E[Y ] = 35 γ = g(x, Y ) + g(λ X, µ Y ) g(λ, µ) d λ = = 1 µ = d 21 = 10 Example 2: X = 30 and Y = 25 Company 1: g(30, 25) = 1 5 = 5 Company 2: g(20, 45) = = 250 Monopoly: g(50, 70) = 200 PGMO 2017 November 14, / 38
29 2 Nodes Network: Theoretical result Theorem When θ, A R such that : λ = µ γ θ = Aθ 1/2 ρ(λ µ) 2 λ µ γ θ = Aθ 3/2 e 2(1 ρ)(λ+µ).θ + o(θ 3/2 e Proof: Calculus via the folded normal distribution. ρ(λ µ)2 2(1 ρ)(λ+µ).θ ) PGMO 2017 November 14, / 38
30 2 Nodes Network: Theoretical result Theorem When θ, A R such that : λ = µ γ θ = Aθ 1/2 ρ(λ µ) 2 λ µ γ θ = Aθ 3/2 e 2(1 ρ)(λ+µ).θ + o(θ 3/2 e Proof: Calculus via the folded normal distribution. Some comments: ρ(λ µ)2 2(1 ρ)(λ+µ).θ ) PGMO 2017 November 14, / 38
31 2 Nodes Network: Theoretical result Theorem When θ, A R such that : λ = µ γ θ = Aθ 1/2 ρ(λ µ) 2 λ µ γ θ = Aθ 3/2 e 2(1 ρ)(λ+µ).θ + o(θ 3/2 e Proof: Calculus via the folded normal distribution. Some comments: ρ(λ µ)2 2(1 ρ)(λ+µ).θ ) Corner points matter: If we keep the same one for all companies (see Ex. 1) then there is no loss. PGMO 2017 November 14, / 38
32 2 Nodes Network: Theoretical result Theorem When θ, A R such that : λ = µ γ θ = Aθ 1/2 ρ(λ µ) 2 λ µ γ θ = Aθ 3/2 e 2(1 ρ)(λ+µ).θ + o(θ 3/2 e Proof: Calculus via the folded normal distribution. Some comments: ρ(λ µ)2 2(1 ρ)(λ+µ).θ ) Corner points matter: If we keep the same one for all companies (see Ex. 1) then there is no loss. We distinguish two regimes: PGMO 2017 November 14, / 38
33 2 Nodes Network: Theoretical result Theorem When θ, A R such that : λ = µ γ θ = Aθ 1/2 ρ(λ µ) 2 λ µ γ θ = Aθ 3/2 e 2(1 ρ)(λ+µ).θ + o(θ 3/2 e Proof: Calculus via the folded normal distribution. Some comments: ρ(λ µ)2 2(1 ρ)(λ+µ).θ ) Corner points matter: If we keep the same one for all companies (see Ex. 1) then there is no loss. We distinguish two regimes: Balanced demands: square root decay. PGMO 2017 November 14, / 38
34 2 Nodes Network: Theoretical result Theorem When θ, A R such that : λ = µ γ θ = Aθ 1/2 ρ(λ µ) 2 λ µ γ θ = Aθ 3/2 e 2(1 ρ)(λ+µ).θ + o(θ 3/2 e Proof: Calculus via the folded normal distribution. Some comments: ρ(λ µ)2 2(1 ρ)(λ+µ).θ ) Corner points matter: If we keep the same one for all companies (see Ex. 1) then there is no loss. We distinguish two regimes: Balanced demands: square root decay. Imbalanced demands: Exponential decay. PGMO 2017 November 14, / 38
35 PoF in General Networks Since the change of corner point induces the loss of efficiency we thus define: C α = { λ s.t. RC(λ) = α λ } PGMO 2017 November 14, / 38
36 PoF in General Networks Since the change of corner point induces the loss of efficiency we thus define: C α = { λ s.t. RC(λ) = α λ } Property α E, C α is such that : It is a closed cone. The intersection of two cones is a plane. Proof: The cones structure comes from the homogeneity of RC. The closedness is due to the fact that C α = (RC(λ) α λ) 1 ({0}). PGMO 2017 November 14, / 38
37 Geometrical intuition λ 2 λ 2 α α β β γ γ λ 1 λ 1 Exponential decay Square root decay PGMO 2017 November 14, / 38
38 Main Theorem: Concepts ρ is not necessarily homogeneous. ρ Λ C α, (1 ρ) Λ C β and Λ C η, i.e. α, β and η are the optimal corner points for respectively company 1, company 2 and the monopoly. The function f used in the main theorem will denote a speed decay which is faster than a square root (i.e. the exponential decay). PGMO 2017 November 14, / 38
39 Main Theorem Theorem Let s suppose that: There is a r.v. ξ such that λ θ = θρ Λ + θσξ with E[ ξ 1 ] < P( ξ > t) = O(f (t)) with f (t) = O(t) Let s define: Then we have: L = α ρ Λ + β (1 ρ) Λ η Λ α C α and β C β γ θ = L + O(f ( θ)) α C α \ C α or β C β \ C β γ θ = L + Θ(θ 1/2 ) PGMO 2017 November 14, / 38
40 Intuition of the proof λ 2 λ 2 α α β β γ γ λ 1 λ 1 Exponential decay Square root decay PGMO 2017 November 14, / 38
41 Interesting cases Applies for both the binomial process and the Poisson process: Binomial process: a R, P( ξ > t) = O(e at2 ) Poisson process: a R, P( ξ > t) = O(e at ) If ρ is homogenous, then α = β = η L = 0 PGMO 2017 November 14, / 38
42 Sufficient condition on the square root regime We know that the square root regime is critical because it decays much slower. But When does it happen? PGMO 2017 November 14, / 38
43 Sufficient condition on the square root regime We know that the square root regime is critical because it decays much slower. But When does it happen? Theorem: Sufficient condition for the square root decay Let s suppose that: The constraint matrix of the dual is not singular E H = {(i, j) x ij > 0} contains at least two connected components. Then the dual optimal point is defined by two different corner points. PGMO 2017 November 14, / 38
44 Sufficient condition on the square root regime We know that the square root regime is critical because it decays much slower. But When does it happen? Theorem: Sufficient condition for the square root decay Let s suppose that: The constraint matrix of the dual is not singular E H = {(i, j) x ij > 0} contains at least two connected components. Then the dual optimal point is defined by two different corner points. Proof: Let s define EĤ = {(i, j) α i α j = τ ij }. Starting from EĤ = E H, we define a dual corner point β which can be slightly modified to saturate new dual constraints and defining another dual corner point. PGMO 2017 November 14, / 38
45 Example: an imbalanced graph with square root decay Λ 1 = [2, 3, 4, 1] Λ 2 = [2, 3, 1, 4] PGMO 2017 November 14, / 38
46 Example: an imbalanced graph with square root decay Λ 1 = [2, 3, 4, 1] Λ 2 = [2, 3, 1, 4] PGMO 2017 November 14, / 38
47 Example: an imbalanced graph with square root decay Λ 1 = [2, 3, 4, 1] Λ 2 = [2, 3, 1, 4] Λ 3 = [2, 3, 3, 2] is optimal for two corner points. PGMO 2017 November 14, / 38
48 Numerical simulations with the previous network Simulations with Λ 1 : Imbalanced demand. Simulations with Λ 3 : Balanced demand. PGMO 2017 November 14, / 38
49 Numerical simulations with real data log(γ θ ) depending on log(θ). TLC Data clustered into 40 stations. PGMO 2017 November 14, / 38
50 Frequency of a square root decay Probability of having two connected components on two months period from TLC dataset. PGMO 2017 November 14, / 38
51 Numerical simulations comparing random processes log(γ θ ) depending on log(θ). Comparison of binomial and Poisson processes. PGMO 2017 November 14, / 38
52 On the necessity of a market maker Those properties allow to draw conclusions on how we can best treat the total demand. We have those three different regimes: PGMO 2017 November 14, / 38
53 On the necessity of a market maker Those properties allow to draw conclusions on how we can best treat the total demand. We have those three different regimes: Highly inhomogenous demand: PoF (θ) L.θ Balanced demand: PoF (θ) θ Imbalanced demand: PoF (θ) e aθ PGMO 2017 November 14, / 38
54 On the necessity of a market maker Those properties allow to draw conclusions on how we can best treat the total demand. We have those three different regimes: Highly inhomogenous demand: PoF (θ) L.θ Balanced demand: PoF (θ) θ Imbalanced demand: PoF (θ) e aθ Phase transition: either lim x PoF (θ) = 0 or + PGMO 2017 November 14, / 38
55 On the necessity of a market maker Those properties allow to draw conclusions on how we can best treat the total demand. We have those three different regimes: Highly inhomogenous demand: PoF (θ) L.θ Balanced demand: PoF (θ) θ Imbalanced demand: PoF (θ) e aθ Phase transition: either lim x PoF (θ) = 0 or + Thus we deduce that to enhance the efficiency we need to: Homogenize demands between companies. Imbalance demands at each node for both companies. PGMO 2017 November 14, / 38
56 Numerical simulations of the two decaying regimes θγ θ depending on θ for the demands Λ 1 and Λ 3 PGMO 2017 November 14, / 38
57 On the necessity of a market maker Though, pricing policies might already aim at balancing demands, thus generating balanced nodes and inducing a square root decay. PGMO 2017 November 14, / 38
58 On the necessity of a market maker Though, pricing policies might already aim at balancing demands, thus generating balanced nodes and inducing a square root decay. But is it possible to transform a balanced demand into two imbalanced one? PGMO 2017 November 14, / 38
59 On the necessity of a market maker Though, pricing policies might already aim at balancing demands, thus generating balanced nodes and inducing a square root decay. But is it possible to transform a balanced demand into two imbalanced one? We have that Λ 1 + Λ 2 = Λ, and they need to keep the same corner point to have at most a square root decay. Assuming the cones are convex, if they have the same corner points then so has the monopoly. Conversely, if the monopoly is optimized on several corner points, thus each company should be optimized on different corner points. PGMO 2017 November 14, / 38
60 On the necessity of a market maker Though, pricing policies might already aim at balancing demands, thus generating balanced nodes and inducing a square root decay. But is it possible to transform a balanced demand into two imbalanced one? We have that Λ 1 + Λ 2 = Λ, and they need to keep the same corner point to have at most a square root decay. Assuming the cones are convex, if they have the same corner points then so has the monopoly. Conversely, if the monopoly is optimized on several corner points, thus each company should be optimized on different corner points. Conclusion: It seems that the demand itself needs to be modified so as to make it imbalanced. PGMO 2017 November 14, / 38
61 Ideas to go further The previous conclusion draws attention on the need of a more complex model, or other approaches: PGMO 2017 November 14, / 38
62 Ideas to go further The previous conclusion draws attention on the need of a more complex model, or other approaches: Queuing theory to study the availability of nodes. On-line Linear Programs to have an unsteady demand. Lumped model instead of clustering the city into stations. Game theoretical models to consider pricing or social welfare. PGMO 2017 November 14, / 38
63 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. PGMO 2017 November 14, / 38
64 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. Those results imply that settling a market maker would be beneficial for the system. PGMO 2017 November 14, / 38
65 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. Those results imply that settling a market maker would be beneficial for the system. Though the models shows some limits and the subject needs further developments. Other questions need to be answered: PGMO 2017 November 14, / 38
66 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. Those results imply that settling a market maker would be beneficial for the system. Though the models shows some limits and the subject needs further developments. Other questions need to be answered: Is there a splitting policy that could dramatically enhance the system s efficiency? PGMO 2017 November 14, / 38
67 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. Those results imply that settling a market maker would be beneficial for the system. Though the models shows some limits and the subject needs further developments. Other questions need to be answered: Is there a splitting policy that could dramatically enhance the system s efficiency? If there is, then could it be feasibly implemented in practice? PGMO 2017 November 14, / 38
68 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. Those results imply that settling a market maker would be beneficial for the system. Though the models shows some limits and the subject needs further developments. Other questions need to be answered: Is there a splitting policy that could dramatically enhance the system s efficiency? If there is, then could it be feasibly implemented in practice? Which economic consequences would it have? Would it preserve the benefits of competition? PGMO 2017 November 14, / 38
69 Conclusion We proved that there were dramatically different behaviours of the loss due to competition. Those results imply that settling a market maker would be beneficial for the system. Though the models shows some limits and the subject needs further developments. Other questions need to be answered: Is there a splitting policy that could dramatically enhance the system s efficiency? If there is, then could it be feasibly implemented in practice? Which economic consequences would it have? Would it preserve the benefits of competition? Thank you! PGMO 2017 November 14, / 38
70 Proof for the exponential decay (1/3) We suppose for sake of simplicity that ρ is homogenous. We define x = ρλ, y = (1 ρ)λ, and δ such that: Where P (α,β) = C α C β. δ = min{min α E d(x, P (α,β)), min α E d(y, P (α,β))} We also denote: η = max η E η Furthermore we have: λ θ C α1 and (θλ λ θ ) C α2 PGMO 2017 November 14, / 38
71 Proof for the exponential decay (2/3) γ θ = θ 1.E[α 1 λθ + α 2 (θλ λθ ) η θλ] = θ 1.E[(ρα 1 + (1 ρ)α 2 η) θλ + (α 1 α 2 ) θσξ] 2 η.( i Λ i ).P( i, ( θσξ) i > θδ) + θ 1 α 1 α 2.E[ θσξ 1.1 { i, ( θσξ)i >θδ} ] 2 η.( i Λ i ).P( i, ( θσξ) i > θδ) + 2θ 1 η θσ 1.E[ ξ 1 i, ( θσξ) i > θδ].p( i, ( θσξ) i > θδ) PGMO 2017 November 14, / 38
72 Proof for the exponential decay (3/3) Thanks to the law of total expectation: E[ ξ 1 i, ( θσξ) i > θδ] < Furthermore: P( i, ( θσξ) i > θδ) P( i, θ σ 1. ξ i > θδ) N 2.P( σ 1. ξ i > δ θ) = O(f ( θ)) PGMO 2017 November 14, / 38
73 Proof for the square root decay (1/2) Demonstration of the lower bound: γ θ = θ 1.E[g(λ θ ) + g(θλ λ θ ) g(θλ)] θ 1.E[(α λ θ + β (θλ λ θ ) β θλ).1 {α σξ β σξ}] + θ 1.E[(β λ θ + α (θλ λ θ ) α θλ).1 {β σξ α σξ}] = θ 1/2.E[(α β) σξ.1 {α σξ β σξ}] + θ 1/2.E[(α β) σξ.1 {β σξ α σξ}] = θ 1/2.E[ (α β) σξ ] = Ω(θ 1/2 ) PGMO 2017 November 14, / 38
74 Proof for the square root decay (2/2) Demonstration of the higher bound: γ θ = θ 1.E[(g(λ θ ) + g(θλ λ θ ) g(θλ))1 { i, ( θσξ)i θδ} ] + θ 1.E[(g(λ θ ) + g(θλ λ θ ) g(θλ))1 { i, ( θσξ)i >θδ} ] θ 1/2.E[ (α β) σξ 1 { i, ( θσξ)i θδ} ] + O(f ( θ)) θ 1/2.E[ (α β) σξ ] + O(f ( θ)) = O(θ 1/2 ) Thus we finally have: γ θ = Θ( θ) PGMO 2017 November 14, / 38
arxiv: v1 [cs.sy] 29 Nov 2017
The Price of Fragmentation in Mobility-on-Demand Services THIBAULT SÉJOURNÉ, Ecole Polytechnique, France SAMITHA SAMARANAYAKE, Cornell University, USA SIDDHARTHA BANERJEE, Cornell University, USA arxiv:1711.1096v1
More informationSession-Based Queueing Systems
Session-Based Queueing Systems Modelling, Simulation, and Approximation Jeroen Horters Supervisor VU: Sandjai Bhulai Executive Summary Companies often offer services that require multiple steps on the
More informationNotes on Continuous Random Variables
Notes on Continuous Random Variables Continuous random variables are random quantities that are measured on a continuous scale. They can usually take on any value over some interval, which distinguishes
More informationCongestion Equilibrium for Differentiated Service Classes Richard T. B. Ma
Congestion Equilibrium for Differentiated Service Classes Richard T. B. Ma School of Computing National University of Singapore Allerton Conference 2011 Outline Characterize Congestion Equilibrium Modeling
More informationLecture Notes 7 Random Processes. Markov Processes Markov Chains. Random Processes
Lecture Notes 7 Random Processes Definition IID Processes Bernoulli Process Binomial Counting Process Interarrival Time Process Markov Processes Markov Chains Classification of States Steady State Probabilities
More informationStein s method and zero bias transformation: Application to CDO pricing
Stein s method and zero bias transformation: Application to CDO pricing ESILV and Ecole Polytechnique Joint work with N. El Karoui Introduction CDO a portfolio credit derivative containing 100 underlying
More informationGrowing competition in electricity industry and the power source structure
Growing competition in electricity industry and the power source structure Hiroaki Ino Institute of Intellectual Property and Toshihiro Matsumura Institute of Social Science, University of Tokyo [Preliminary
More informationElementary Probability. Exam Number 38119
Elementary Probability Exam Number 38119 2 1. Introduction Consider any experiment whose result is unknown, for example throwing a coin, the daily number of customers in a supermarket or the duration of
More information1 Functions and Graphs
1 Functions and Graphs 1.1 Functions Cartesian Coordinate System A Cartesian or rectangular coordinate system is formed by the intersection of a horizontal real number line, usually called the x axis,
More informationMATH The Derivative as a Function - Section 3.2. The derivative of f is the function. f x h f x. f x lim
MATH 90 - The Derivative as a Function - Section 3.2 The derivative of f is the function f x lim h 0 f x h f x h for all x for which the limit exists. The notation f x is read "f prime of x". Note that
More informationConvex polytopes, interacting particles, spin glasses, and finance
Convex polytopes, interacting particles, spin glasses, and finance Sourav Chatterjee (UCB) joint work with Soumik Pal (Cornell) Interacting particles Systems of particles governed by joint stochastic differential
More informationLecture 4: Bernoulli Process
Lecture 4: Bernoulli Process Hyang-Won Lee Dept. of Internet & Multimedia Eng. Konkuk University Lecture 4 Hyang-Won Lee 1 / 12 Stochastic Process A stochastic process is a mathematical model of a probabilistic
More information1/2 1/2 1/4 1/4 8 1/2 1/2 1/2 1/2 8 1/2 6 P =
/ 7 8 / / / /4 4 5 / /4 / 8 / 6 P = 0 0 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Andrei Andreevich Markov (856 9) In Example. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 P (n) = 0
More informationResiliency of Mobility-as-a-Service Systems to Denial-of-Service Attacks
Resiliency of Mobility-as-a-Service Systems to Denial-of-Service Attacks Jérôme Thai 1 and Chenyang Yuan 1 and Alexandre M. Bayen 1,2 Abstract Mobility-as-a-Service (MaaS) systems such as ride sharing
More informationRANDOM SIMULATIONS OF BRAESS S PARADOX
RANDOM SIMULATIONS OF BRAESS S PARADOX PETER CHOTRAS APPROVED: Dr. Dieter Armbruster, Director........................................................ Dr. Nicolas Lanchier, Second Committee Member......................................
More information2 Random Variable Generation
2 Random Variable Generation Most Monte Carlo computations require, as a starting point, a sequence of i.i.d. random variables with given marginal distribution. We describe here some of the basic methods
More information2905 Queueing Theory and Simulation PART IV: SIMULATION
2905 Queueing Theory and Simulation PART IV: SIMULATION 22 Random Numbers A fundamental step in a simulation study is the generation of random numbers, where a random number represents the value of a random
More informationSupply chains models
Supply chains models Maria Ivanova Technische Universität München, Department of Mathematics, Haupt Seminar 2016 Garching, Germany July 3, 2016 Overview Introduction What is Supply chain? Traffic flow
More informationChapter 2: Random Variables
ECE54: Stochastic Signals and Systems Fall 28 Lecture 2 - September 3, 28 Dr. Salim El Rouayheb Scribe: Peiwen Tian, Lu Liu, Ghadir Ayache Chapter 2: Random Variables Example. Tossing a fair coin twice:
More informationCalculus Review Session. Brian Prest Duke University Nicholas School of the Environment August 18, 2017
Calculus Review Session Brian Prest Duke University Nicholas School of the Environment August 18, 2017 Topics to be covered 1. Functions and Continuity 2. Solving Systems of Equations 3. Derivatives (one
More informationThe Growth of Functions. A Practical Introduction with as Little Theory as possible
The Growth of Functions A Practical Introduction with as Little Theory as possible Complexity of Algorithms (1) Before we talk about the growth of functions and the concept of order, let s discuss why
More informationLecture 12 The Spatial Price Equilibrium Problem
Lecture 12 The Spatial Price Equilibrium Problem Dr. Anna Nagurney John F. Smith Memorial Professor Isenberg School of Management University of Massachusetts Amherst, Massachusetts 01003 c 2009 Parallel
More informationPERCOLATION IN SELFISH SOCIAL NETWORKS
PERCOLATION IN SELFISH SOCIAL NETWORKS Charles Bordenave UC Berkeley joint work with David Aldous (UC Berkeley) PARADIGM OF SOCIAL NETWORKS OBJECTIVE - Understand better the dynamics of social behaviors
More informationLecture 7: Lagrangian Relaxation and Duality Theory
Lecture 7: Lagrangian Relaxation and Duality Theory (3 units) Outline Lagrangian dual for linear IP Lagrangian dual for general IP Dual Search Lagrangian decomposition 1 / 23 Joseph Louis Lagrange Joseph
More informationRecap. Probability, stochastic processes, Markov chains. ELEC-C7210 Modeling and analysis of communication networks
Recap Probability, stochastic processes, Markov chains ELEC-C7210 Modeling and analysis of communication networks 1 Recap: Probability theory important distributions Discrete distributions Geometric distribution
More informationMath 3215 Intro. Probability & Statistics Summer 14. Homework 5: Due 7/3/14
Math 325 Intro. Probability & Statistics Summer Homework 5: Due 7/3/. Let X and Y be continuous random variables with joint/marginal p.d.f. s f(x, y) 2, x y, f (x) 2( x), x, f 2 (y) 2y, y. Find the conditional
More informationStochastic Models of Manufacturing Systems
Stochastic Models of Manufacturing Systems Ivo Adan Organization 2/47 7 lectures (lecture of May 12 is canceled) Studyguide available (with notes, slides, assignments, references), see http://www.win.tue.nl/
More informationRandom Geometric Graphs
Random Geometric Graphs Mathew D. Penrose University of Bath, UK Networks: stochastic models for populations and epidemics ICMS, Edinburgh September 2011 1 MODELS of RANDOM GRAPHS Erdos-Renyi G(n, p):
More informationGreenwich Public Schools Mathematics Curriculum Objectives. Calculus
Mathematics Curriculum Objectives Calculus June 30, 2006 NUMERICAL AND PROPORTIONAL REASONING Quantitative relationships can be expressed numerically in multiple ways in order to make connections and simplify
More informationEcon Slides from Lecture 10
Econ 205 Sobel Econ 205 - Slides from Lecture 10 Joel Sobel September 2, 2010 Example Find the tangent plane to {x x 1 x 2 x 2 3 = 6} R3 at x = (2, 5, 2). If you let f (x) = x 1 x 2 x3 2, then this is
More informationTEST CODE: MMA (Objective type) 2015 SYLLABUS
TEST CODE: MMA (Objective type) 2015 SYLLABUS Analytical Reasoning Algebra Arithmetic, geometric and harmonic progression. Continued fractions. Elementary combinatorics: Permutations and combinations,
More informationAsymptotic Statistics-III. Changliang Zou
Asymptotic Statistics-III Changliang Zou The multivariate central limit theorem Theorem (Multivariate CLT for iid case) Let X i be iid random p-vectors with mean µ and and covariance matrix Σ. Then n (
More informationOptimization. A first course on mathematics for economists
Optimization. A first course on mathematics for economists Xavier Martinez-Giralt Universitat Autònoma de Barcelona xavier.martinez.giralt@uab.eu II.3 Static optimization - Non-Linear programming OPT p.1/45
More informationFall 2009 Math 113 Final Exam Solutions. f(x) = 1 + ex 1 e x?
. What are the domain and range of the function Fall 9 Math 3 Final Exam Solutions f(x) = + ex e x? Answer: The function is well-defined everywhere except when the denominator is zero, which happens when
More informationInterlude: Practice Final
8 POISSON PROCESS 08 Interlude: Practice Final This practice exam covers the material from the chapters 9 through 8. Give yourself 0 minutes to solve the six problems, which you may assume have equal point
More informationIntroduction and Overview STAT 421, SP Course Instructor
Introduction and Overview STAT 421, SP 212 Prof. Prem K. Goel Mon, Wed, Fri 3:3PM 4:48PM Postle Hall 118 Course Instructor Prof. Goel, Prem E mail: goel.1@osu.edu Office: CH 24C (Cockins Hall) Phone: 614
More informationSchumpeterian Growth Models
Schumpeterian Growth Models Yin-Chi Wang The Chinese University of Hong Kong November, 2012 References: Acemoglu (2009) ch14 Introduction Most process innovations either increase the quality of an existing
More informationA New Random Graph Model with Self-Optimizing Nodes: Connectivity and Diameter
A New Random Graph Model with Self-Optimizing Nodes: Connectivity and Diameter Richard J. La and Maya Kabkab Abstract We introduce a new random graph model. In our model, n, n 2, vertices choose a subset
More informationHW1 solutions. 1. α Ef(x) β, where Ef(x) is the expected value of f(x), i.e., Ef(x) = n. i=1 p if(a i ). (The function f : R R is given.
HW1 solutions Exercise 1 (Some sets of probability distributions.) Let x be a real-valued random variable with Prob(x = a i ) = p i, i = 1,..., n, where a 1 < a 2 < < a n. Of course p R n lies in the standard
More informationError Functions & Linear Regression (1)
Error Functions & Linear Regression (1) John Kelleher & Brian Mac Namee Machine Learning @ DIT Overview 1 Introduction Overview 2 Univariate Linear Regression Linear Regression Analytical Solution Gradient
More informationSolution Set for Homework #1
CS 683 Spring 07 Learning, Games, and Electronic Markets Solution Set for Homework #1 1. Suppose x and y are real numbers and x > y. Prove that e x > ex e y x y > e y. Solution: Let f(s = e s. By the mean
More informationSimulation of stationary processes. Timo Tiihonen
Simulation of stationary processes Timo Tiihonen 2014 Tactical aspects of simulation Simulation has always a goal. How to organize simulation to reach the goal sufficiently well and without unneccessary
More informationA Note on Cost Reducing Alliances in Vertically Differentiated Oligopoly. Abstract
A Note on Cost Reducing Alliances in Vertically Differentiated Oligopoly Frédéric DEROÏAN FORUM Abstract In a vertically differentiated oligopoly, firms raise cost reducing alliances before competing with
More informationMachine Learning. Support Vector Machines. Fabio Vandin November 20, 2017
Machine Learning Support Vector Machines Fabio Vandin November 20, 2017 1 Classification and Margin Consider a classification problem with two classes: instance set X = R d label set Y = { 1, 1}. Training
More information6.207/14.15: Networks Lecture 3: Erdös-Renyi graphs and Branching processes
6.207/14.15: Networks Lecture 3: Erdös-Renyi graphs and Branching processes Daron Acemoglu and Asu Ozdaglar MIT September 16, 2009 1 Outline Erdös-Renyi random graph model Branching processes Phase transitions
More informationSpatial Economics and Potential Games
Outline Spatial Economics and Potential Games Daisuke Oyama Graduate School of Economics, Hitotsubashi University Hitotsubashi Game Theory Workshop 2007 Session Potential Games March 4, 2007 Potential
More informationCoarsening process in the 2d voter model
Alessandro Tartaglia (LPTHE) Coarsening in the 2d voter model May 8, 2015 1 / 34 Coarsening process in the 2d voter model Alessandro Tartaglia LPTHE, Université Pierre et Marie Curie alessandro.tartaglia91@gmail.com
More informationStat 516, Homework 1
Stat 516, Homework 1 Due date: October 7 1. Consider an urn with n distinct balls numbered 1,..., n. We sample balls from the urn with replacement. Let N be the number of draws until we encounter a ball
More informationStochastic Models in Computer Science A Tutorial
Stochastic Models in Computer Science A Tutorial Dr. Snehanshu Saha Department of Computer Science PESIT BSC, Bengaluru WCI 2015 - August 10 to August 13 1 Introduction 2 Random Variable 3 Introduction
More informationThe Geography of Development: Evaluating Migration Restrictions and Coastal Flooding
The Geography of Development: Evaluating Migration Restrictions and Coastal Flooding Klaus Desmet SMU Dávid Krisztián Nagy Princeton University Esteban Rossi-Hansberg Princeton University World Bank, February
More informationProbability Models. 4. What is the definition of the expectation of a discrete random variable?
1 Probability Models The list of questions below is provided in order to help you to prepare for the test and exam. It reflects only the theoretical part of the course. You should expect the questions
More informationECE 302 Division 1 MWF 10:30-11:20 (Prof. Pollak) Final Exam Solutions, 5/3/2004. Please read the instructions carefully before proceeding.
NAME: ECE 302 Division MWF 0:30-:20 (Prof. Pollak) Final Exam Solutions, 5/3/2004. Please read the instructions carefully before proceeding. If you are not in Prof. Pollak s section, you may not take this
More informationModelling Czech and Slovak labour markets: A DSGE model with labour frictions
Modelling Czech and Slovak labour markets: A DSGE model with labour frictions Daniel Němec Faculty of Economics and Administrations Masaryk University Brno, Czech Republic nemecd@econ.muni.cz ESF MU (Brno)
More informationExpectation. DS GA 1002 Probability and Statistics for Data Science. Carlos Fernandez-Granda
Expectation DS GA 1002 Probability and Statistics for Data Science http://www.cims.nyu.edu/~cfgranda/pages/dsga1002_fall17 Carlos Fernandez-Granda Aim Describe random variables with a few numbers: mean,
More informationh Edition Money in Search Equilibrium
In the Name of God Sharif University of Technology Graduate School of Management and Economics Money in Search Equilibrium Diamond (1984) Navid Raeesi Spring 2014 Page 1 Introduction: Markets with Search
More informationCDA6530: Performance Models of Computers and Networks. Chapter 3: Review of Practical Stochastic Processes
CDA6530: Performance Models of Computers and Networks Chapter 3: Review of Practical Stochastic Processes Definition Stochastic process X = {X(t), t2 T} is a collection of random variables (rvs); one rv
More informationAnalysis of Engineering and Scientific Data. Semester
Analysis of Engineering and Scientific Data Semester 1 2019 Sabrina Streipert s.streipert@uq.edu.au Example: Draw a random number from the interval of real numbers [1, 3]. Let X represent the number. Each
More informationOD-Matrix Estimation using Stable Dynamic Model
OD-Matrix Estimation using Stable Dynamic Model Yuriy Dorn (Junior researcher) State University Higher School of Economics and PreMoLab MIPT Alexander Gasnikov State University Higher School of Economics
More informationR&D Collaboration in Collusive Networks
R&D Collaboration in Collusive Networks Gizem Korkmaz European University Institute December, 2011 Market Sharing Agreements In recent years, the scope for explicit or implicit market sharing agreements
More informationCollaborative Network Formation in Spatial Oligopolies
Collaborative Network Formation in Spatial Oligopolies 1 Shaun Lichter, Terry Friesz, and Christopher Griffin arxiv:1108.4114v1 [math.oc] 20 Aug 2011 Abstract Recently, it has been shown that networks
More informationRobust Predictions in Games with Incomplete Information
Robust Predictions in Games with Incomplete Information joint with Stephen Morris (Princeton University) November 2010 Payoff Environment in games with incomplete information, the agents are uncertain
More informationMoment-based Availability Prediction for Bike-Sharing Systems
Moment-based Availability Prediction for Bike-Sharing Systems Jane Hillston Joint work with Cheng Feng and Daniël Reijsbergen LFCS, School of Informatics, University of Edinburgh http://www.quanticol.eu
More informationThe strictly 1/2-stable example
The strictly 1/2-stable example 1 Direct approach: building a Lévy pure jump process on R Bert Fristedt provided key mathematical facts for this example. A pure jump Lévy process X is a Lévy process such
More informationStat 426 : Homework 1.
Stat 426 : Homework 1. Moulinath Banerjee University of Michigan Announcement: The homework carries 120 points and contributes 10 points to the total grade. (1) A geometric random variable W takes values
More informationA model for competitive markets
A model for competitive markets Manuela Girotti MATH 345 Differential Equations Contents 1 Competitive markets 1 1.1 Competitive market with only one good......................... 2 1.2 Competitive market
More informationCOMP9334 Capacity Planning for Computer Systems and Networks
COMP9334 Capacity Planning for Computer Systems and Networks Week 2: Operational Analysis and Workload Characterisation COMP9334 1 Last lecture Modelling of computer systems using Queueing Networks Open
More informationu xx + u yy = 0. (5.1)
Chapter 5 Laplace Equation The following equation is called Laplace equation in two independent variables x, y: The non-homogeneous problem u xx + u yy =. (5.1) u xx + u yy = F, (5.) where F is a function
More informationDemand and Trip Prediction in Bike Share Systems
Demand and Trip Prediction in Bike Share Systems Team members: Zhaonan Qu SUNet ID: zhaonanq December 16, 2017 1 Abstract 2 Introduction Bike Share systems are becoming increasingly popular in urban areas.
More informationBertrand Model of Price Competition. Advanced Microeconomic Theory 1
Bertrand Model of Price Competition Advanced Microeconomic Theory 1 ҧ Bertrand Model of Price Competition Consider: An industry with two firms, 1 and 2, selling a homogeneous product Firms face market
More information56 CHAPTER 3. POLYNOMIAL FUNCTIONS
56 CHAPTER 3. POLYNOMIAL FUNCTIONS Chapter 4 Rational functions and inequalities 4.1 Rational functions Textbook section 4.7 4.1.1 Basic rational functions and asymptotes As a first step towards understanding
More informationELE539A: Optimization of Communication Systems Lecture 16: Pareto Optimization and Nonconvex Optimization
ELE539A: Optimization of Communication Systems Lecture 16: Pareto Optimization and Nonconvex Optimization Professor M. Chiang Electrical Engineering Department, Princeton University March 16, 2007 Lecture
More informationContinuous RVs. 1. Suppose a random variable X has the following probability density function: π, zero otherwise. f ( x ) = sin x, 0 < x < 2
STAT 4 Exam I Continuous RVs Fall 7 Practice. Suppose a random variable X has the following probability density function: f ( x ) = sin x, < x < π, zero otherwise. a) Find P ( X < 4 π ). b) Find µ = E
More informationRevenue Maximization in a Cloud Federation
Revenue Maximization in a Cloud Federation Makhlouf Hadji and Djamal Zeghlache September 14th, 2015 IRT SystemX/ Telecom SudParis Makhlouf Hadji Outline of the presentation 01 Introduction 02 03 04 05
More informationLearning Spectral Graph Segmentation
Learning Spectral Graph Segmentation AISTATS 2005 Timothée Cour Jianbo Shi Nicolas Gogin Computer and Information Science Department University of Pennsylvania Computer Science Ecole Polytechnique Graph-based
More informationLinear Programming Methods
Chapter 11 Linear Programming Methods 1 In this chapter we consider the linear programming approach to dynamic programming. First, Bellman s equation can be reformulated as a linear program whose solution
More informationModeling Rare Events
Modeling Rare Events Chapter 4 Lecture 15 Yiren Ding Shanghai Qibao Dwight High School April 24, 2016 Yiren Ding Modeling Rare Events 1 / 48 Outline 1 The Poisson Process Three Properties Stronger Property
More informationFunctions of Several Variables
Functions of Several Variables Partial Derivatives Philippe B Laval KSU March 21, 2012 Philippe B Laval (KSU) Functions of Several Variables March 21, 2012 1 / 19 Introduction In this section we extend
More informationFundamental Characteristics of Urban Transportation Services
Fundamental Characteristics of Urban Transportation Services Anton J. Kleywegt School of Industrial and Systems Engineering Georgia Institute of Technology Smart Urban Transportation Forum Institute for
More informationOptimization 4. GAME THEORY
Optimization GAME THEORY DPK Easter Term Saddle points of two-person zero-sum games We consider a game with two players Player I can choose one of m strategies, indexed by i =,, m and Player II can choose
More informationA Study on Performance Analysis of Queuing System with Multiple Heterogeneous Servers
UNIVERSITY OF OKLAHOMA GENERAL EXAM REPORT A Study on Performance Analysis of Queuing System with Multiple Heterogeneous Servers Prepared by HUSNU SANER NARMAN husnu@ou.edu based on the papers 1) F. S.
More informationContents. Set Theory. Functions and its Applications CHAPTER 1 CHAPTER 2. Preface... (v)
(vii) Preface... (v) CHAPTER 1 Set Theory Definition of Set... 1 Roster, Tabular or Enumeration Form... 1 Set builder Form... 2 Union of Set... 5 Intersection of Sets... 9 Distributive Laws of Unions and
More informationManaging Call Centers with Many Strategic Agents
Managing Call Centers with Many Strategic Agents Rouba Ibrahim, Kenan Arifoglu Management Science and Innovation, UCL YEQT Workshop - Eindhoven November 2014 Work Schedule Flexibility 86SwofwthewqBestwCompanieswtowWorkwForqwofferw
More informationMEI STRUCTURED MATHEMATICS CONCEPTS FOR ADVANCED MATHEMATICS, C2. Practice Paper C2-C
MEI Mathematics in Education and Industry MEI STRUCTURED MATHEMATICS CONCEPTS FOR ADVANCED MATHEMATICS, C Practice Paper C-C Additional materials: Answer booklet/paper Graph paper MEI Examination formulae
More information6. Qualitative Solutions of the TISE
6. Qualitative Solutions of the TISE Copyright c 2015 2016, Daniel V. Schroeder Our goal for the next few lessons is to solve the time-independent Schrödinger equation (TISE) for a variety of one-dimensional
More informationUsing Economic Contexts to Advance in Mathematics
Using Economic Contexts to Advance in Mathematics Erik Balder University of Utrecht, Netherlands DEE 2013 presentation, Exeter Erik Balder (Mathematical Institute, University of Utrecht)using economic
More informationEECS 126 Probability and Random Processes University of California, Berkeley: Fall 2014 Kannan Ramchandran November 13, 2014.
EECS 126 Probability and Random Processes University of California, Berkeley: Fall 2014 Kannan Ramchandran November 13, 2014 Midterm Exam 2 Last name First name SID Rules. DO NOT open the exam until instructed
More informationELE539A: Optimization of Communication Systems Lecture 6: Quadratic Programming, Geometric Programming, and Applications
ELE539A: Optimization of Communication Systems Lecture 6: Quadratic Programming, Geometric Programming, and Applications Professor M. Chiang Electrical Engineering Department, Princeton University February
More informationPart I Stochastic variables and Markov chains
Part I Stochastic variables and Markov chains Random variables describe the behaviour of a phenomenon independent of any specific sample space Distribution function (cdf, cumulative distribution function)
More informationLimits, Continuity, and the Derivative
Unit #2 : Limits, Continuity, and the Derivative Goals: Study and define continuity Review limits Introduce the derivative as the limit of a difference quotient Discuss the derivative as a rate of change
More informationLinear Model Predictive Control for Queueing Networks in Manufacturing and Road Traffic
Linear Model Predictive Control for ueueing Networks in Manufacturing and Road Traffic Yoni Nazarathy Swinburne University of Technology, Melbourne. Joint work with: Erjen Lefeber (manufacturing), Hai
More information1 Overview. 2 Extreme Points. AM 221: Advanced Optimization Spring 2016
AM 22: Advanced Optimization Spring 206 Prof. Yaron Singer Lecture 7 February 7th Overview In the previous lectures we saw applications of duality to game theory and later to learning theory. In this lecture
More information6.2 Their Derivatives
Exponential Functions and 6.2 Their Derivatives Copyright Cengage Learning. All rights reserved. Exponential Functions and Their Derivatives The function f(x) = 2 x is called an exponential function because
More informationStochastic Proximal Gradient Algorithm
Stochastic Institut Mines-Télécom / Telecom ParisTech / Laboratoire Traitement et Communication de l Information Joint work with: Y. Atchade, Ann Arbor, USA, G. Fort LTCI/Télécom Paristech and the kind
More information10/22/16. 1 Math HL - Santowski SKILLS REVIEW. Lesson 15 Graphs of Rational Functions. Lesson Objectives. (A) Rational Functions
Lesson 15 Graphs of Rational Functions SKILLS REVIEW! Use function composition to prove that the following two funtions are inverses of each other. 2x 3 f(x) = g(x) = 5 2 x 1 1 2 Lesson Objectives! The
More informationData Abundance and Asset Price Informativeness. On-Line Appendix
Data Abundance and Asset Price Informativeness On-Line Appendix Jérôme Dugast Thierry Foucault August 30, 07 This note is the on-line appendix for Data Abundance and Asset Price Informativeness. It contains
More informationLecture 4: Two-point Sampling, Coupon Collector s problem
Randomized Algorithms Lecture 4: Two-point Sampling, Coupon Collector s problem Sotiris Nikoletseas Associate Professor CEID - ETY Course 2013-2014 Sotiris Nikoletseas, Associate Professor Randomized Algorithms
More informationAPPM 1235 Exam 2 Spring 2018
APPM 1235 Exam 2 Spring 2018 INSTRUCTIONS: Outside paper and electronic devices are not permitted. Exam is worth 100 points. Neatness counts. Unless indicated, answers with no supporting work may receive
More informationBranch-and-cut (and-price) for the chance constrained vehicle routing problem
Branch-and-cut (and-price) for the chance constrained vehicle routing problem Ricardo Fukasawa Department of Combinatorics & Optimization University of Waterloo May 25th, 2016 ColGen 2016 joint work with
More informationDistributed Optimization. Song Chong EE, KAIST
Distributed Optimization Song Chong EE, KAIST songchong@kaist.edu Dynamic Programming for Path Planning A path-planning problem consists of a weighted directed graph with a set of n nodes N, directed links
More informatione x3 dx dy. 0 y x 2, 0 x 1.
Problem 1. Evaluate by changing the order of integration y e x3 dx dy. Solution:We change the order of integration over the region y x 1. We find and x e x3 dy dx = y x, x 1. x e x3 dx = 1 x=1 3 ex3 x=
More information