Browsing by Subject "Learning"

Humans make decisions in highly complex physical, economic and social environments. In order to adaptively choose, the human brain has to learn about- and attend to- sensory cues that provide information about the potential outcome of different courses of action. Here I present three event-related potential (ERP) studies, in which I evaluated the role of the interactions between attention and reward learning in economic decision-making. I focused my analyses on three ERP components (Chap. 1): (1) the N2pc, an early lateralized ERP response reflecting the lateralized focus of visual; (2) the feedback-related negativity (FRN), which reflects the process by which the brain extracts utility from feedback; and (3) the P300 (P3), which reflects the amount of attention devoted to feedback-processing. I found that learned stimulus-reward associations can influence the rapid allocation of attention (N2pc) towards outcome-predicting cues, and that differences in this attention allocation process are associated with individual differences in economic decision performance (Chap. 2). Such individual differences were also linked to differences in neural responses reflecting the amount of attention devoted to processing monetary outcomes (P3) (Chap. 3). Finally, the relative amount of attention devoted to processing rewards for oneself versus others (as reflected by the P3) predicted both charitable giving and self-reported engagement in real-life altruistic behaviors across individuals (Chap. 4). Overall, these findings indicate that attention and reward processing interact and can influence each other in the brain. Moreover, they indicate that individual differences in economic choice behavior are associated both with biases in the manner in which attention is drawn towards sensory cues that inform subsequent choices, and with biases in the way that attention is allocated to learn from the outcomes of recent choices.

Children learn crucial skills like speech by imitating the behavior of skilled adults. Similarly, juvenile zebra finches learn to sing by learning to imitate adults. This song learning process enables laboratory study of juvenile imitative learning. But it also poses behavioral quantification challenges. Zebra finches produce hundreds of thousands of complex vocalizations during vocal development. These undergo learned changes with respect to acoustic features that are relevant to the animal but experimentally unknown \textit{a priori}. Recent developments in machine learning provide tools to reduce the dimensionality of complex behaviors, plausibly simplifying this inference challenge. These tools have not been validated on or applied to song learning problems.

Here, I validate the use of an autoencoder to extract copying-relevant features from zebra finch song. Then, I develop tools to quantify developmental song change with respect to extracted features. In particular, I generate forward models that quantify developmental changes in syllable acoustic distributions. I also develop a method to score syllable maturity on a rendition-by-rendition basis. Both these techniques reveal circadian behavioral patterns that differ between normally developing and untutored juveniles, suggesting that tutoring not only sets target song acoustics; it directly affects intrinsic features of practice behavior. Critically, these tools enable making concrete predictions from otherwise abstract song learning theories.