Data science pecha kucha

[2014-10-30 Thu 19:15] speakers: MIT SENSEable city labs event: Data Science SG ** Agent based transport modelling from fare data speaker: Pieter Fourie and Sergio Ordonez *** extract model parameters from fare data transport demand transport supply behavior parameters *** run simulations in MATSim study effects of network redesign bus route modification ** Understanding metropolitian patterns of daily encounters speaker: Lijun Sun Stanley Milgram’s concept of familiar strangers study fare data, in the same bus at least twice in study period -> familiar stanger ** My thoughts on Big Data speaker: Daniel Dahlmeier

• volume, velocity, variety
• the turkey problem
• spurious correlations
• no data like more data (Banko and Brill, 2001) ** Mining geo spatial data for location-based services speaker: Naveen Nandan reducing the storage requirements for point sequence data using a grid, dwelling duration, transition ** Call details record analysis speaker: Tracey Li information about cell tower, time, who you called

limited by spatial and temporal resolution. location of caller approximated by location of cell tower.

strategic location and opening times for retail outlets

how is your social network affected by the population of your city? senseable city lab “urban villages” project ** A new data visualization platform speaker: Oliver Senn

datacollider.io