Lectures

LECTURE 6

1. Why are humans better at understanding patterns than processes?

processes - need to be solved serially - step by step. However, neuron speed is too slow - only 200 calculation per second. Still 100 bilions of neurons are working at same time. This parallelism allow brain to perform feats of pattern recognition - remember faces, create metaphore. Human brain relies on precomputing its analyses and store them for future reference.

How do systems store information?

In self-organising structure, retrieving them, creating clusters of similar information. Existence of data management - as a product of collective afterthougt.

How do systems learn?

Learning is not only of being aware of information - it is also about storing information and knowing where to find it.

- being able to recognize and response to changing patterns.

They "learn" by the possibility of altering system behaviour

How do systems adapt?

By changing behaviour. It keeps usefulls structures and modify unusefull structures. "Useful structures" were selected by the forces of bological or cultural evolution.

Personally,I think that agents are able to learn = to change their previous behaviour and that the system is able to adapt = to change its usual responses to be ore corresponding to actually changing conditions.

What are the opposing forces that “keep the drift and tumult of history at bay”?

Kind of self-organising stickiness, defined by usefulness of the elements/clusters. Existence of positive feedbacks which reinforce the presence of usefull clusters (-> clustering = self-perpetuating system) and eliminate presence of waste elements (by negative feedback). This one can move away or change their behaviour to be more in accordance with their neighbors. If the structure is staleble over-time it means that is usefull and in accordance with its neigbours.

***********************************************************************************************************************

Homework for Lecture 3

1. Open the “Segregation” model and explore how the parameter settings influence the emerging output

neighbourhood is defined as 8 cells/patches surrounding the turtle
calculation of the % of hapiness (30% of hapiness means that 30% of all neighbours of red turtle should be the red)
if the turtle has 0 neighbours - it is equally happy
proximity of turtles depend on %-similar-wanted, but not necessary, one turtle shoul be happy if it is not surrounded by any turtle by any color - how to change it in code?

***********************************************************************************************************************

LECTURE 2

Complexity and City

Manchester - formed by paterns of industrias zones and inhabitant zones. Industrial zones are linking the rivers (to Liverpool), main roads and freeways as is the case actually. Areas are intensively builded-up, extent mostly in the southern part of Manchester. Inhabitants buildings are relatively small in comparisons to industrial area. Because of small possibilities of daily traveling, the most inhabitants would build up their houses the nearest possibly to industrial park - close to their employement. The prices increased as the amount off inhabitants. Building are small and sqeezed in the town. Industrical areas are well places, extent and planned to correspond the market demand.

who are the top-down agents in the planning of cities?

architect, investors, city planners, zoning laws, planning commissions...

who are the bottom-up agents in forcing emerging patterns?

citizens, streets, environment, surface...

Which agents are autonomous and how do they exhibit their individuality?

environment - exists and could be modified (surfacing)
inhabitants - own preferences about living, organising their personal areas (garden, house, ...)

By what rules do agents interact?

feedback - tolerance/repugnance between neighbours and neighbourhood, cyclic process of positive or negative feed back - am I satistief with my environment? If not, could I change it or shouldI move away?
depending on time/space/scale

What major feedbacks are present?

does my emplacement work? Am I satisfied or not? Demands are changing over time

How are non-linearities present?

How can adaptation take place?

evaluation over time and space

How does self-organization occur?

By 'what' rules do agents interact? (road from Condon hall to Duck store) - in 'NetLogo code'

agents - men, cars, bikes
move - any direction but can only go forward
aim - others side of the road
rules -
- min distance 80 cm betweem two persons
- person - car - 3 m
- person - bike - 1 m
- vision - 60 degrees
- move - jump to nearly patch,
chack neigbhourdhood - move when it is free

***********************************************************************************************************************