Zhenzhou Wu

• McGill University

Bag-of-Words with TF-IDF or other weighting schemes is commonly adopted ways for document representation. However, they fail to capture sequential or semantic information in the sentence, and would lead to high-dimensional vector due to misspelling, acronyms and so on. Distributed word embedding and even document embedding methods are proposed to e...

Traditionally, classifying large hierarchical labels with more than 10000 distinct traces can only be achieved with flatten labels. Although flatten labels is feasible, it misses the hierarchical information in the labels. Hierarchical models like HSVM by \cite{vural2004hierarchical} becomes impossible to train because of the sheer number of SVMs i...

Despite the success of deep learning on many fronts especially image and speech, its application in text classification often is still not as good as a simple linear SVM on n-gram TF-IDF representation especially for smaller datasets. Deep learning tends to emphasize on sentence level semantics when learning a representation with models like recurr...

Deep Neural Networks (DNN) have been successful in en- hancing noisy speech signals. Enhancement is achieved by learning a nonlinear mapping function from the features of the corrupted speech signal to that of the reference clean speech signal. The quality of predicted features can be improved by providing additional side channel information that i...

In our study, we investigate the effectiveness of different models to the purchasing behaviour at YOOCHOOSE website. This paper provide a direct method in modeling the buying pattern in a clicking session by simply using the time-stamp of the clicks and show that the result is comparable to using more massive feature engineering that requires sessi...

This paper proposes a deep denoising auto-encoder technique to extract better acoustic features for speech synthesis. The technique allows us to automatically extract low-dimensional features from high dimensional spectral features in a non-linear, data-driven, unsupervised way. We compared the new stochastic feature extractor with conventional mel...

In this paper we present the techniques used for the University of Montréal's team submissions to the 2013 Emotion Recognition in the Wild Challenge. The challenge is to classify the emotions expressed by the primary human subject in short video clips extracted from feature length movies. This involves the analysis of video clips of acted scenes la...