Investors (and the intelligent agent) ascribe an intrinsic value to stability. By using the mapping from Figure 9 and small multiples to draw each month separately, every field snapshot is a moment in time, and the apparent flow mobility shows progression in portfolio positions. Therefore, one can easily identify which flow structures tend to remain unchanged, and when a major event occurred; the streamlined plot (Figure 9.d) is ideal for such type of visual analysis. The vector plot (Figure 9.c) presents the “intensity” dimension lacking in the streams, which are focused on directionality. The LIC (line integral convolution) rendering (Figure 9.c) is a human friendly and aesthetically pleasant format that helps reinforce the flow structure without losing analytical capability, especially if one makes it overlap the vector map, or the streamlines, depending on the data dimension to analyze (by using LIC the entropy of the visualization increases, more information is conveyed). The mesh network representation of the flow (Figure 9.d) is a machine format that maintains some of the human friendly features of the other visualizations –one can easily see how computation by clusters might become a useful device to reveal equilibrium points in the vector field topology: unstable nodes (sources or saddles), stable focus (spiral sink), stable centers, etc. This representation is obtained by drawing mesh divisions between every line or polygon generated by a plotting function, in this case, the one obtained after the stream plot. Finally, given the duality between singularities in vector fields and network structure, the fields analysis is suitable to be implemented at scale.
Applications
Intelligence services
Although privacy coins offer desirable features in some settings, for national security purposes is often needed to understand the user's clusters at least at a macro level. Trust asymmetry (in the blockchain / off-chain boundaries) reveals information such as geography and audiences, which allows to re-construct digital personae (groups' identity profiles); it also provides insights into hidden states and phase transitions within those multiplex networks. The applications include mapping the character of Zcash and Monero communities and providing partial source/destination metadata related to the use of secret contracts and similar technologies that facilitate coin mixing (which obscure the original source of cryptocurrency used within the protocol). These applications are also relevant to market intelligence, a practice that becomes more challenging as international privacy standards strengthen.
Financial Risk prediction
Ensemble prediction with evolutionary computing augments both technical analysis and time series ARNN results by providing context and by providing explicit trustability ranges. The intelligent agent is thus capable of reasoning about interventions: what if I change X? This is the alternative to creating static models -- rather, we create models dynamically and perform spatiotemporal encoding of information using a blockchain as a substrate for the intelligence. This also implies a sort of "value by memory" where a system that has experienced more about the world is more valuable than one that consists only of the algorithm.
Political risk prediction
The political economy is a contest for attention modulated by reputation, and thus a perfect fit for context-aware AI.
In multiples areas, from influencing public opinion to political campaign financing, demand is influenced by the available supply (of information).