It is increasingly common to see machine learning techniques applied in conjunction with computational modeling for data-driven research in neuroscience. Such applications include using machine learning for model development, particularly for optimization of parameters based on electrophysiological constraints. Alternatively, machine learning can be used to validate and enhance techniques for experimental data analysis or to analyze model simulation data in large-scale modeling studies, which is the approach I apply here. I use simulations of biophysically-realistic cortical neuron models to supplement a common feature-based technique for analysis of electrophysiological signals. I leverage these simulated electrophysiological signals to perform feature selection that provides an improved method for neuron-type classification. Additionally, I validate an unsupervised approach that extends this improved feature selection to discover signatures associated with neuron morphologies - performing in vivo histology in effect. The result is a simulation-based discovery of the underlying synaptic conditions responsible for patterns of extracellular signatures that can be applied to understand both simulation and experimental data. I also use unsupervised learning techniques to identify common channel mechanisms underlying electrophysiological behaviors of cortical neuron models. This work relies on an open-source database containing a large number of computational models for cortical neurons. I perform a quantitative data-driven analysis of these previously published ion channel and neuron models that uses information shared across models as opposed to information limited to individual models. The result is simulation-based discovery of model sub-types at two spatial scales which map functional relationships between activation/inactivation properties of channel family model sub-types to electrophysiological properties of cortical neuron model sub-types. Further, the combination of unsupervised learning techniques and parameter visualizations serve to integrate characterizations of model electrophysiological behavior across scales.

Neuron models that behave like their biological counterparts are essential for computational neuroscience.Reduced neuron models, which abstract away biological mechanisms in the interest of speed and interpretability, have received much attention due to their utility in large scale simulations of the brain, but little care has been taken to ensure that these models exhibit behaviors that closely resemble real neurons.
In order to improve the verisimilitude of these reduced neuron models, I developed an optimizer that uses genetic algorithms to align model behaviors with those observed in experiments.
I verified that this optimizer was able to recover model parameters given only observed physiological data; however, I also found that reduced models nonetheless had limited ability to reproduce all observed behaviors, and that this varied by cell type and desired behavior.
These challenges can partly be surmounted by carefully designing the set of physiological features that guide the optimization. In summary, we found evidence that reduced neuron model optimization had the potential to produce reduced neuron models for only a limited range of neuron types.

Speech intelligibility measures how much a speaker can be understood by a listener. Traditional measures of intelligibility, such as word accuracy, are not sufficient to reveal the reasons of intelligibility degradation. This dissertation investigates the underlying sources of intelligibility degradations from both perspectives of the speaker and the listener. Segmental phoneme errors and suprasegmental lexical boundary errors are developed to reveal the perceptual strategies of the listener. A comprehensive set of automated acoustic measures are developed to quantify variations in the acoustic signal from three perceptual aspects, including articulation, prosody, and vocal quality. The developed measures have been validated on a dysarthric speech dataset with various severity degrees. Multiple regression analysis is employed to show the developed measures could predict perceptual ratings reliably. The relationship between the acoustic measures and the listening errors is investigated to show the interaction between speech production and perception. The hypothesize is that the segmental phoneme errors are mainly caused by the imprecise articulation, while the sprasegmental lexical boundary errors are due to the unreliable phonemic information as well as the abnormal rhythm and prosody patterns. To test the hypothesis, within-speaker variations are simulated in different speaking modes. Significant changes have been detected in both the acoustic signals and the listening errors. Results of the regression analysis support the hypothesis by showing that changes in the articulation-related acoustic features are important in predicting changes in listening phoneme errors, while changes in both of the articulation- and prosody-related features are important in predicting changes in lexical boundary errors. Moreover, significant correlation has been achieved in the cross-validation experiment, which indicates that it is possible to predict intelligibility variations from acoustic signal.

The marmoset monkey (Callithrix jacchus) is a new-world primate species native to South America rainforests. Because they rely on vocal communication to navigate and survive, marmosets have evolved as a promising primate model to study vocal production, perception, cognition, and social interactions. The purpose of this project is to provide an initial assessment on the vocal repertoire of a marmoset colony raised at Arizona State University and call types they use in different social conditions. The vocal production of a colony of 16 marmoset monkeys was recorded in 3 different conditions with three repeats of each condition. The positive condition involves a caretaker distributing food, the negative condition involves an experimenter taking a marmoset out of his cage to a different room, and the control condition is the normal state of the colony with no human interference. A total of 5396 samples of calls were collected during a total of 256 minutes of audio recordings. Call types were analyzed in semi-automated computer programs developed in the Laboratory of Auditory Computation and Neurophysiology. A total of 5 major call types were identified and their variants in different social conditions were analyzed. The results showed that the total number of calls and the type of calls made differed in the three social conditions, suggesting that monkey vocalization signals and depends on the social context.

Much evidence has shown that first language (L1) plays an important role in the formation of L2 phonological system during second language (L2) learning process. This combines with the fact that different L1s have distinct phonological patterns to indicate the diverse L2 speech learning outcomes for speakers from different L1 backgrounds. This dissertation hypothesizes that phonological distances between accented speech and speakers' L1 speech are also correlated with perceived accentedness, and the correlations are negative for some phonological properties. Moreover, contrastive phonological distinctions between L1s and L2 will manifest themselves in the accented speech produced by speaker from these L1s. To test the hypotheses, this study comes up with a computational model to analyze the accented speech properties in both segmental (short-term speech measurements on short-segment or phoneme level) and suprasegmental (long-term speech measurements on word, long-segment, or sentence level) feature space. The benefit of using a computational model is that it enables quantitative analysis of L1's effect on accent in terms of different phonological properties. The core parts of this computational model are feature extraction schemes to extract pronunciation and prosody representation of accented speech based on existing techniques in speech processing field. Correlation analysis on both segmental and suprasegmental feature space is conducted to look into the relationship between acoustic measurements related to L1s and perceived accentedness across several L1s. Multiple regression analysis is employed to investigate how the L1's effect impacts the perception of foreign accent, and how accented speech produced by speakers from different L1s behaves distinctly on segmental and suprasegmental feature spaces. Results unveil the potential application of the methodology in this study to provide quantitative analysis of accented speech, and extend current studies in L2 speech learning theory to large scale. Practically, this study further shows that the computational model proposed in this study can benefit automatic accentedness evaluation system by adding features related to speakers' L1s.

Auditory scene analysis (ASA) is the process through which listeners parse and organize their acoustic environment into relevant auditory objects. ASA functions by exploiting natural regularities in the structure of auditory information. The current study investigates spectral envelope and its contribution to the perception of changes in pitch and loudness. Experiment 1 constructs a perceptual continuum of twelve f0- and intensity-matched vowel phonemes (i.e. a pure timbre manipulation) and reveals spectral envelope as a primary organizational dimension. The extremes of this dimension are i (as in “bee”) and Ʌ (“bun”). Experiment 2 measures the strength of the relationship between produced f0 and the previously observed phonetic-pitch continuum at three different levels of phonemic constraint. Scat performances and, to a lesser extent, recorded interviews were found to exhibit changes in accordance with the natural regularity; specifically, f0 changes were correlated with the phoneme pitch-height continuum. The more constrained case of lyrical singing did not exhibit the natural regularity. Experiment 3 investigates participant ratings of pitch and loudness as stimuli vary in f0, intensity, and the phonetic-pitch continuum. Psychophysical functions derived from the results reveal that moving from i to Ʌ is equivalent to a .38 semitone decrease in f0 and a .75 dB decrease in intensity. Experiment 4 examines the potentially functional aspect of the pitch, loudness, and spectral envelope relationship. Detection thresholds of stimuli in which all three dimensions change congruently (f0 increase, intensity increase, Ʌ to i) or incongruently (no f0 change, intensity increase, i to Ʌ) are compared using an objective version of the method of limits. Congruent changes did not provide a detection benefit over incongruent changes; however, when the contribution of phoneme change was removed, congruent changes did offer a slight detection benefit, as in previous research. While this relationship does not offer a detection benefit at threshold, there is a natural regularity for humans to produce phonemes at higher f0s according to their relative position on the pitch height continuum. Likewise, humans have a bias to detect pitch and loudness changes in phoneme sweeps in accordance with the natural regularity.

Music training is associated with measurable physiologic changes in the auditory pathway. Benefits of music training have also been demonstrated in the areas of working memory, auditory attention, and speech perception in noise. The purpose of this study was to determine whether long-term auditory experience secondary to music training enhances the ability to detect, learn, and recall new words.

Participants consisted of 20 young adult musicians and 20 age-matched non-musicians. In addition to completing word recognition and non-word detection tasks, each participant learned 10 nonsense words in a rapid word-learning task. All tasks were completed in quiet and in multi-talker babble. Next-day retention of the learned words was examined in isolation and in context. Cortical auditory evoked responses to vowel stimuli were recorded to obtain latencies and amplitudes for the N1, P2, and P3a components. Performance was compared across groups and listening conditions. Correlations between the behavioral tasks and the cortical auditory evoked responses were also examined.

No differences were found between groups (musicians vs. non-musicians) on any of the behavioral tasks. Nor did the groups differ in cortical auditory evoked response latencies or amplitudes, with the exception of P2 latencies, which were significantly longer in musicians than in non-musicians. Performance was significantly poorer in babble than in quiet on word recognition and non-word detection, but not on word learning, learned-word retention, or learned-word detection. CAEP latencies collapsed across group were significantly longer and amplitudes were significantly smaller in babble than in quiet. P2 latencies in quiet were positively correlated with word recognition in quiet, while P3a latencies in babble were positively correlated with word recognition and learned-word detection in babble. No other significant correlations were observed between CAEPs and performance on behavioral tasks.

These results indicated that, for young normal-hearing adults, auditory experience resulting from long-term music training did not provide an advantage for learning new information in either favorable (quiet) or unfavorable (babble) listening conditions. Results of the present study suggest that the relationship between music training and the strength of cortical auditory evoked responses may be more complex or too weak to be observed in this population.

To localize different sound sources in an environment, the auditory system analyzes acoustic properties of sounds reaching the ears to determine the exact location of a sound source. Successful sound localization is important for improving signal detection and speech intelligibility in a noisy environment. Sound localization is not a uni-sensory experience, and can be influenced by visual information (e.g., the ventriloquist effect). Vision provides contexts and organizes the auditory space for the auditory system. This investigation tracks eye movement of human subjects using a non-invasive eye-tracking system and evaluates the impact of visual stimulation on localization of a phantom sound source generated through timing-based stereophony. It was hypothesized that gaze movement could reveal the way in which visual stimulation (LED lights) shifts the perception of a sound source. However, the results show that subjects do not always move their gaze towards the light direction even when they experience strong visual capture. On average, the gaze direction indicates the perceived sound location with and without light stimulation.

Hearing and vision are two senses that most individuals use on a daily basis. The simultaneous presentation of competing visual and auditory stimuli often affects our sensory perception. It is often believed that vision is the more dominant sense over audition in spatial localization tasks. Recent work suggests that visual information can influence auditory localization when the sound is emanating from a physical location or from a phantom location generated through stereophony (the so-called "summing localization"). The present study investigates the role of cross-modal fusion in an auditory localization task. The focuses of the experiments are two-fold: (1) reveal the extent of fusion between auditory and visual stimuli and (2) investigate how fusion is correlated with the amount of visual bias a subject experiences. We found that fusion often occurs when light flash and "summing localization" stimuli were presented from the same hemifield. However, little correlation was observed between the magnitude of visual bias and the extent of perceived fusion between light and sound stimuli. In some cases, subjects reported distinctive locations for light and sound and still experienced visual capture.

Head turning is a common sound localization strategy in primates. A novel system that can track head movement and acoustic signals received at the entrance to the ear canal was tested to obtain binaural sound localization information during fast head movement of marmoset monkey. Analysis of binaural information was conducted with a focus on inter-aural level difference (ILD) and inter-aural time difference (ITD) at various head positions over time. The results showed that during fast head turns, the ITDs showed significant and clear changes in trajectory in response to low frequency stimuli. However, significant phase ambiguity occurred at frequencies greater than 2 kHz. Analysis of ITD and ILD information with animal vocalization as the stimulus was also tested. The results indicated that ILDs may provide more information in understanding the dynamics of head movement in response to animal vocalizations in the environment. The primary significance of this experimentation is the successful implementation of a system capable of simultaneously recording head movement and acoustic signals at the ear canals. The collected data provides insight into the usefulness of ITD and ILD as binaural cues during head movement.