Alexandr Pshenichkin

Learn More
We present reliability solutions for adaptable network RAM systems running on general-purpose clusters. Network RAM allows nodes with over-committed memory to swap pages over the network, storing them in the idle RAM of other nodes and avoiding swapping to slow, local disk. An adaptable network RAM system adjusts the amount of RAM currently available for(More)
Intelligent Adaptive Curiosity (IAC) is an attempt at motivation-based robot learning. IAC has very limited ability to predict and categorize time-dependent elements of its environment, however. In this paper, we attempt to integrate IAC's strong points into a more capable learning algorithm. We implement this by adding a simple recurrent network as a(More)
Bridges and similar non-obstructing features inhibit correct flow routing on high-resolution digital elevation models because their apparent elevation does not reflect the elevation at which water may pass underneath them. Our goal is to identify such features using the elevation data so that flow-routing algorithms may find paths under them correctly. We(More)
  • 1