cmn 150 practice actual exam questions and correct answers|

Allow us to test hypothesis. How it is special? or are they just random? Digital footprint-- Collect Data-- analyze-- Hypothesis= check on this network or To deduce our hypothesis. - correct answers Random Networks Random Graph - correct answers Erdos-Renyi Graph Erdos-Renyi Graph/Random Graph - correct answers -Start with n nodes -Form independent links (with a certain p (probability) OR from M(number) of links -Properties of the graph (what is the average degree of the network [ numerical solution] or analytical solution)) -We get 'almost sure' properties with specific G(n,...) setups ( 1)what is the degree distribution of the network? 2)What is the likelihood of getting a hub with almost all links concentrated on it alone? 3) What are most nodes connected in one giant component?) Likelihood of getting a property depends on P (probability), but !Non-linear phase transitions emerges! at... - correct answers Tipping points is a connected component of a given random graph that contains a finite fraction of the entire graph's. Groups can merge and create this component. Tipping point= Threshold Function 1 connection p/node = we can all be connected= emergence of groups - correct answers Giant Component Scale-free networks - correct answers -Add nodes with uniform likelihood Start with # nods fully connected Add new nodes with # links to existing nodes with equal likelihood -Add nodes with non-uniform likelihood Some nodes have many more links than others Preferential attachment growing through preferential attachment= add nodes - correct answers How Scale-free networks evolve? likelihood is proportional to number of existing degrees ( if you are more popular you are more likely to get more connections) (hub) exponentially Few have a lot of connections/ and a lot nodes who have 1 degree) - correct answers Preferential attachment random+ preferential attachment - most of our friends are connected to highly connected people -geographical proximity - preferential attachment- exponentially few has exponentially many/ and vice versa. - correct answers Hybrid Models Network structures are not random - correct answers Small World networks - High level of clustering ( is unlikely/slower to reach everybody) - small average path length:likely to quickly reach all others on average l - correct answers characteristics of small world networks High clustering and small average path length between clusters - correct answers Sweet Spot of Small World Networks -2 degrees of separation - benefits/costs are direct = # of links between nodes Benefits*2= indirect benefits the most stable and beneficial network. Shorter path length are more valuable than longer path lengths - correct answers Hub & Spoke/ star networks Equal distribution of benefits and costs among the nodes - correct answers Efficient Network Social Efficiency - sum of value of everybody is high Social Stability- nobody cant get better off by changing the network -Does not always work need to give without taking away from some or have to take away without taking away from others - correct answers Strategic Network Intervention Diffuse faster, because of the hubs - correct answers Preferential network Diffusion in Netlogo Slower than Scale-free diffusion= tipping points - correct answers Random Graph Diffusion in NL hubs are important*2 - more likely get and pass innovations/infections/ etc - Pagerak centrality ( hub) allows to stop diffusion. - correct answers Hub Diffusion in NL - Abstract or physical structure that can potentially represent real-world phenomena - correct answers Model 3 ABmodel movement rules 1. Separation 2. Alignment 3. Cohesion - correct answers Craig Reynold's "Boids" Cultural groups are internally similar but maintain differences with others Individual Assumptions: -Homophily -Influence Features Dymanics Use Random Grapgh - correct answers Axelrod's cultural model -different maps and models can be both right, representing different aspects-- sub-aspects - correct answers Modelling vs. Mapping Coding- computer language - correct answers What do we use for ABM - Total is different than sum of parts (racists lead to mixed societies) - Dependence on initial conditions - Phase transition ( be aware of modeling assumptions) - Invariant Distribution of Equilibria -Micro Motives lead to micro-social behavior - Reality is more complex - correct answers Schelling's segregation model Simulates population with limited spatially-distributed resources available -tick is a period in time -patch is space -darker yellow more sugar - correct answers Sugarscape Agents -Traits - Rules Environment -DONUT/ Sugar Lattice -Traits -Rules - correct answers Sugarscape Components Agents Traits and Rules Environment Traits and Rules - correct answers Sugarscape Components: Fixed Agents Traits Environment Traits - correct answers Sugarscape Components: Variable Constant grow back of sugar is gradual -population went down/ do not find stable point - agents move to better patches ( more sugary patches) - vision increased/ metabolism decreased -more evolutionary pressure ( depends from sugar grow back) - correct answers Sugarscape 2 Grow back of sugar instantaneous - population decreased- stable - vision increased/ metabolism decreased - correct answers Sugarscape 1 effective - correct answers Great vision in Sugarscape efficient/productive - correct answers Great metabolism in Sugarscape wealth - correct answers Sugar in Sugarscape Inequality (Lorenz curve) agents die (starvation or age)/constant population growth Pareto-law - correct answers Sugarscape 3