We won't see Google's bizarre boondoggle make an appearance at this year's conference, and we didn't really need AllThingsD's reported confirmation of that fact to predict its absence, but it's nice to have it.
Angeli and Manning [3], for example, propose to use the relations extracted from texts to enlarge databases of known facts and to predict facts, introducing the notion of fact similarity.
Thus, demographic characteristics (relative abundance and distribution) of potential hosts were in fact shown to predict the evolution of specialized races or species of parasites [ 3].
It is noteworthy that the type of pathology alone was in fact sufficient to predict the response (CML patients had globally better response scores than GIST patients, P<0.001).
As such, there is a need to explore the ability of the FACT-G to predict individual dimensions of the EQ-5D and SF-6D; the predicted dimensions could be used to calculate an overall utility score.
The poor ability of the FACT-G to predict EQ-5D utilities may be influenced by the brevity of the preference-based instrument (i.e., three levels to define the five dimensions).
For random sampling, however, it is in fact straightforward to predict the number of interactions: if a fraction p of nodes has been sampled, then the fraction of edges that has been sampled is simply the fraction of pairs of nodes, i.e. a random subnet with sampling probability p will have a proportion of p2 of the edges.
At any rate, the real sea change has already taken place: The government has gone from investigating crimes after the fact to predicting who is going to commit them in the future, as if we're entering the world of the sci-fi film Minority Report.
About 20 years ago economists decided to abandon data and go for an ever more abstract series of models that are logically consistent but not tested against facts, and unable to predict real crises.
But these so-called "stylized facts" are supposed to predict human behavior; they don't necessarily pretend to mirror it.
