The performance of deep learning frameworks could be significantly improved through considering the particular underlying structures for each dataset. In this thesis, I summarize our three work about boosting the performance of deep learning models through leveraging structures of the data. In the first work, we theoretically justify that, for...
DNA microarray technology is a powerful tool for analyzing patterns in gene expression data for thousands of genes. Due to a number of systematic variations in microarray experiments, the raw gene expression data is often obfuscated by undesirable technical noises. Various normalization techniques were designed in an attempt to remove...
This dissertation presents nonlinear terahertz (THz) properties of carbon nanomaterials investigated by time-resolved high-field THz spectroscopy. In order to determine THz characteristics of nanomaterials, we performed THz power spectrum measurement, THz raster imaging, THz time-domain spectroscopy (THz-TDS) and time-resolved pump-probe experiment on two different types of single layer graphene and...
The advent of deep learning models leads to a substantial improvement in a wide range of NLP tasks, achieving state-of-art performances without any hand-crafted features. However, training deep models requires a massive amount of labeled data. Labeling new data as a new task or domain emerges consumes time and efforts...
Learning to recognize objects is a fundamental and essential step in human perception and understanding of the world. Accordingly, research of object discovery across diverse modalities plays a pivotal role in the context of computer vision. This field not only contributes significantly to enhancing our understanding of visual information but...
Messenger RNA (mRNA) therapy has massive potential to treat various genetic disorders by providing effective proteins to the diseased cells and tissues. Its therapeutic applications include protein replacement, genome editing, and immunotherapy. For successful mRNA treatment, mRNA must be delivered to the cytosol of target cells. One of the leading...
The potential for nanomedicine to reduce toxicity and improve therapeutic response to radiation is vast. Clinically, this potential has thus far only begun to be realized with the recent success of the radioenhancing nanoparticle, NBTXR3 by Nanobiotix. Radioenhancing nanoparticles, however, represent only a small class of nanomaterials that may be...
This dissertation presents novel, field-activated smart material systems for the actuation and control of autonomous robots. Smart materials, a type of material whose properties can be changed with an external stimuli, represent a promising direction to expand upon existing robotic control and actuation methods, particularly in the sub-fields of soft...