A YouTube video sent to Apple blog 9to5Mac lets you hear the difference between the old Siri and the new one. In the video, a more lifelike Siri says certain words much more naturally, as if they were recorded by an actual person. The digital assistant's cadence has also improved, as the pitch of Siri's voice drops at the end of clauses just like a natural English speaker's would. Siri's voice has undergone periodic tweaks and fixes, but none so major as when Apple (AAPL, Tech30) released iOS 7 in 2013. In that update, Siri got an entirely new sound, switching from actress Susan Bennett's voice to two other unnamed actors (one male and one female).

Active learning has shown to reduce the number of experiments needed to obtain high-confidence drug-target predictions. However, in order to actually save experiments using active learning, it is crucial to have a method to evaluate the quality of the current prediction and decide when to stop the experimentation process. Only by applying reliable stoping criteria to active learning, time and costs in the experimental process can be actually saved. We compute active learning traces on simulated drug-target matrices in order to learn a regression model for the accuracy of the active learner. By analyzing the performance of the regression model on simulated data, we design stopping criteria for previously unseen experimental matrices. We demonstrate on four previously characterized drug effect data sets that applying the stopping criteria can result in upto 40% savings of the total experiments for highly accurate predictions.

We show that if $F$ is a convex class of functions that is $L$-subgaussian, the error rate of learning problems generated by independent noise is equivalent to a fixed point determined by `local' covering estimates of the class, rather than by the gaussian averages. To that end, we establish new sharp upper and lower estimates on the error rate for such problems.

Many modern tools in machine learning and signal processing, such as sparse dictionary learning, principal component analysis (PCA), non-negative matrix factorization (NMF), $K$-means clustering, etc., rely on the factorization of a matrix obtained by concatenating high-dimensional vectors from a training collection. While the idealized task would be to optimize the expected quality of the factors over the underlying distribution of training vectors, it is achieved in practice by minimizing an empirical average over the considered collection. The focus of this paper is to provide sample complexity estimates to uniformly control how much the empirical average deviates from the expected cost function. Standard arguments imply that the performance of the empirical predictor also exhibit such guarantees. The level of genericity of the approach encompasses several possible constraints on the factors (tensor product structure, shift-invariance, sparsity \ldots), thus providing a unified perspective on the sample complexity of several widely used matrix factorization schemes. The derived generalization bounds behave proportional to $\sqrt{\log(n)/n}$ w.r.t.\ the number of samples $n$ for the considered matrix factorization techniques.

Matrix completion has attracted significant recent attention in many fields including statistics, applied mathematics and electrical engineering. Current literature on matrix completion focuses primarily on independent sampling models under which the individual observed entries are sampled independently. Motivated by applications in genomic data integration, we propose a new framework of structured matrix completion (SMC) to treat structured missingness by design. Specifically, our proposed method aims at efficient matrix recovery when a subset of the rows and columns of an approximately low-rank matrix are observed. We provide theoretical justification for the proposed SMC method and derive lower bound for the estimation errors, which together establish the optimal rate of recovery over certain classes of approximately low-rank matrices. Simulation studies show that the method performs well in finite sample under a variety of configurations. The method is applied to integrate several ovarian cancer genomic studies with different extent of genomic measurements, which enables us to construct more accurate prediction rules for ovarian cancer survival.

Protein contacts contain important information for protein structure and functional study, but contact prediction from sequence remains very challenging. Both evolutionary coupling (EC) analysis and supervised machine learning methods are developed to predict contacts, making use of different types of information, respectively. This paper presents a group graphical lasso (GGL) method for contact prediction that integrates joint multi-family EC analysis and supervised learning. Different from existing single-family EC analysis that uses residue co-evolution information in only the target protein family, our joint EC analysis uses residue co-evolution in both the target family and its related families, which may have divergent sequences but similar folds. To implement joint EC analysis, we model a set of related protein families using Gaussian graphical models (GGM) and then co-estimate their precision matrices by maximum-likelihood, subject to the constraint that the precision matrices shall share similar residue co-evolution patterns. To further improve the accuracy of the estimated precision matrices, we employ a supervised learning method to predict contact probability from a variety of evolutionary and non-evolutionary information and then incorporate the predicted probability as prior into our GGL framework. Experiments show that our method can predict contacts much more accurately than existing methods, and that our method performs better on both conserved and family-specific contacts.

Deep Neural Network (DNN) acoustic models have yielded many state-of-the-art results in Automatic Speech Recognition (ASR) tasks. More recently, Recurrent Neural Network (RNN) models have been shown to outperform DNNs counterparts. However, state-of-the-art DNN and RNN models tend to be impractical to deploy on embedded systems with limited computational capacity. Traditionally, the approach for embedded platforms is to either train a small DNN directly, or to train a small DNN that learns the output distribution of a large DNN. In this paper, we utilize a state-of-the-art RNN to transfer knowledge to small DNN. We use the RNN model to generate soft alignments and minimize the Kullback-Leibler divergence against the small DNN. The small DNN trained on the soft RNN alignments achieved a 3.93 WER on the Wall Street Journal (WSJ) eval92 task compared to a baseline 4.54 WER or more than 13% relative improvement.

Twitter Idioms are one of the important types of hashtags that spread in Twitter. In this paper, we propose a classifier that can stratify the Idioms from the other kind of hashtags with 86.93% accuracy and high precision and recall rate. We then learn regression models on the stratified samples (Idioms and non-Idioms) separately to predict the popularity of the Idioms. This stratification not only itself allows us to make more accurate predictions but also makes it possible to include Idiom-specific features to separately improve the accuracy for the Idioms. Experimental results show that such stratification during the training phase followed by inclusion of Idiom-specific features leads to an overall improvement of 11.13% and 19.56% in correlation coefficient over the baseline method after the 7th and the 11th month respectively.

Guilds, a primary form of community in many online games, are thought to both aid gameplay and act as social entities. This work uses a three-year scrape of one game, World of Warcraft, to study the relationship between guild membership and advancement in the game as measured by character leveling, a defining and often studied metric. 509 guilds and 90,581 characters are included in the analysis from a three-year period with over 36 million observations, with linear regression to measure the effect of guild membership. Overall findings indicate that guild membership does not aid character leveling to any significant extent. The benefits of guilds may be replicated by players in smaller guilds or not in guilds through game affordances and human sociability.

The proliferation of mobile handheld devices in combination with the technological advancements in mobile computing has led to a number of innovative services that make use of the location information available on such devices. Traditional yellow pages websites have now moved to mobile platforms, giving the opportunity to local businesses and potential, near-by, customers to connect. These platforms can offer an affordable advertisement channel to local businesses. One of the mechanisms offered by location-based social networks (LBSNs) allows businesses to provide special offers to their customers that connect through the platform. We collect a large time-series dataset from approximately 14 million venues on Foursquare and analyze the performance of such campaigns using randomization techniquesand (non-parametric) hypothesis testing with statistical bootstrapping. Our main finding indicates that this type of promotions are not as effective as anecdote success stories might suggest. Finally, we design classifiers by extracting three different types of features that are able to provide an educated decision on whether a special offer campaign for a local business will succeed or not both in short and long term.