Faculty

Speaker: Daniel Recoskie, PhD candidate

We propose a new method for learning filters for the 2D discrete wavelet transform. We extend our previous work on the 1D wavelet transform in order to process images. We show that the 2D wavelet transform can be represented as a modified convolutional neural network (CNN). Doing so allows us to learn wavelet filters from data by gradient descent. Our learned wavelets are similar to traditional wavelets which are typically derived using Fourier methods.

Speaker: Amir-Hossein Karimi, Master’s candidate

The story of this work is dimensionality reduction. Dimensionality reduction is a method that takes as input a point-set P of n points in \(R^d\) where d is typically large and attempts to find a lower-dimensional representation of that dataset, in order to ease the burden of processing for down-stream algorithms. In today’s landscape of machine learning, researchers and practitioners work with datasets that either have a very large number of samples and/or include high-dimensional samples. Therefore, dimensionality reduction is applied as a pre-processing technique primarily to overcome the curse of dimensionality.

Speaker: Zhucheng Tu, Master's Candidate

Modelling the similarity of two sentences is an important problem in natural language processing and information retrieval, with applications in tasks such as paraphrase identification and answer selection in question answering. The Multi-Perspective Convolutional Neural Network (MP-CNN) is a model that improved previous state-of-the-art models in 2015 and has remained a popular model for sentence similarity tasks. However, until now, there has not been a rigorous study of how the model actually achieves competitive accuracy. 

Speaker: Areej Alhothali, PhD Candidate

We propose an extension to graph-based sentiment lexicon induction methods by incorporating distributed and semantic word representations in building the similarity graph to expand a three-dimensional sentiment lexicon. We also present a comprehensive evaluation of four graph-based propagation approaches using different word representations and similarity metrics.

Speaker: Priyank Jaini, PhD

We consider the problem of inferring a sequence of hidden states associated with a sequence of observations produced by an individual within a population. Instead of learning a single sequence model for the population (which does not account for variations within the population), we learn a set of basis sequence models based on different individuals.

Speaker: Aaron R. Voelker, PhD Candidate

One of the central challenges in neuroscience involves understanding how dynamic stimuli can be processed by neural mechanisms to drive behavior. By synthesizing models from neuroscience with tools from signal processing and control theory, we derive the feedback weights required to represent a low-frequency input signal by delaying it along a low-dimensional manifold, using a recurrently connected network of biologically plausible spiking neurons. This is shown to nonlinearly encode the continuous-time history of an input signal by optimally compressing it into a low-dimensional subspace.

Speaker: Milad Khaki, PhD Candidate

The project involves the analysis of procurement auctions for water infrastructure maintenance projects for Cities of Niagara Falls and London. The primary concern of this proposal is to generate a reliable mechanism for importing different formats of the tender summaries, provided as input data, of the projects and unify them in a database for further exploration and processing.

Speaker: Zhengkun Shang, Master's Candidate

Emotions are an essential part of human social interactions. By integrating an automatic affect recognizer into an artificial system, the system can detect humans’ emotions and provide personal responses. We aim to build a prompting system that uses a virtual human with emotional interaction capabilities to help persons with a cognitive disability to complete daily activities independently.

Speaker: Ivana Kajic, PhD Candidate

The semantic fluency task has been used to understand the effects of semantic relationships on human memory search. A variety of computational models have been proposed that explain human behavioral data, yet it remains unclear how millions of spiking neurons work in unison to realize the cognitive processes involved in memory search.

Speaker: Areej Alhothali, PhD Candidate

We propose an affect control theory-based (ACT) model to predict the temporal progression of the interactants' emotions and their optimal behavior over a sequence of interactions. ACT has a solid foundation in sociology to interpret, understand, and predict human social interactions. In this study, we extend the sociological mathematical model in ACT by learning and incorporating parameters from real data.