We propose different approaches to infer causal influences between agents in a network using only observed time series. This includes graphical models to depict causal relationships in the network, algorithms to identify ...

Cognitive developmental robotics unites machine learning and neuroscience with the aim of creating robots which display the robustness and efficiency of human cognition. This document proposes a novel method for acquiring ...

Decision-making is based on the integration of sensory information, motivational states and memory, from which assessment of outcomes and optimal action selection emerge. In foraging, animals tend to make decisions that ...

This work examines an information theoretic quantity known as directed information, which measures statistically causal influences between processes. It is shown to be a general quantity, applicable to arbitrary probability ...

In this thesis, we look closely at two fundamental problems that arise within the context of multimedia blind watermark decoding and timing channels steganalysis. The central problem considered, loosely speaking, is that ...

In many sensor network applications, one is interested only in computing some relevant \textit{function} of the sensor measurements. In this thesis, we study optimal strategies for in-network computation and communication ...

We propose a new method for learning policies for large, partially observable Markov decision processes (POMDPs) that require long time horizons for planning. Computing optimal policies for POMDPs is an intractable problem ...

Performance in most complex cognitive and psychomotor tasks improves with training, yet the extent of improvement varies among individuals. Is it possible to forecast the benefit that a person might reap from training? ...

Brain-computer interfaces (BCI) allow human subjects to interact with exogenous systems through neural signals. The goal of this interaction may be to induce behaviors or properties in either the exogenous system or ...