Generative Adversarial Networks (GANs) have been successful in producing outstanding results in areas as diverse as image, video, and text generation. Building on these successes, a large number of empirical studies have validated the benefits of the cousin approach called Wasserstein GANs (WGANs), which brings stabilization in the training process. In the present paper, we add a new stone to the edifice by proposing some theoretical advances in the properties of WGANs. First, we properly define the architecture of WGANs in the context of integral probability metrics parameterized by neural networks and highlight some of their basic mathematical features. We stress in particular interesting optimization properties arising from the use of a parametric 1-Lipschitz discriminator. Then, in a statistically-driven approach, we study the convergence of empirical WGANs as the sample size tends to infinity, and clarify the adversarial effects of the generator and the discrimi-nator by underlining some trade-off properties. These features are finally illustrated with experiments using both synthetic and real-world datasets.

We study sample complexity of optimizing "hill-climbing friendly" functions defined on a graph under noisy observations. We define a notion of convexity, and we show that a variant of best-arm identification can find a near-optimal solution after a small number of queries that is independent of the size of the graph. For functions that have local minima and are nearly convex, we show a sample complexity for the classical simulated annealing under noisy observations. We show effectiveness of the greedy algorithm with restarts and the simulated annealing on problems of graph-based nearest neighbor classification as well as a web document re-ranking application.

We modeled the Quora question pairs dataset to identify a similar question. The dataset that we use is provided by Quora. The task is a binary classification. We tried several methods and algorithms and different approach from previous works. For feature extraction, we used Bag of Words including Count Vectorizer, and Term Frequency-Inverse Document Frequency with unigram for XGBoost and CatBoost. Furthermore, we also experimented with WordPiece tokenizer which improves the model performance significantly. We achieved up to 97 percent accuracy. Code and Dataset.

And how to improve UMAP. This is the thirteenth article of my column Mathematical Statistics and Machine Learning for Life Sciences where I try to explain some mysterious analytical techniques used in Bioinformatics, Biomedicine, Genetics etc. in a simple way. In the previous post How Exactly UMAP works I started with an intuitive explanation of the math behind UMAP. The best way to learn it is to program UMAP from scratch, this is what we are going to do today. The idea of this post is to show that it is relatively easy for everyone to create their own neighbor graph dimension reduction technique that can provide even better visualization than UMAP. It is going to be lots of coding, buckle up!

Topological Data Analysis (TDA) provides novel approaches that allow us to analyze the geometrical shapes and topological structures of a dataset. As one important application, TDA can be used for data visualization and dimension reduction. We follow the framework of circular coordinate representation, which allows us to perform dimension reduction and visualization for high-dimensional datasets on a torus using persistent cohomology. In this paper, we propose a method to adapt the circular coordinate framework to take into account sparsity in high-dimensional applications. We use a generalized penalty function instead of an $L_{2}$ penalty in the traditional circular coordinate algorithm. We provide simulation experiments and real data analysis to support our claim that circular coordinates with generalized penalty will accommodate the sparsity in high-dimensional datasets under different sampling schemes while preserving the topological structures.

It can be seen The differential geometric approach to probability theory as a natural choice of metric as it is the only Riemannian and statistics has met increasing interest in the past metric that is invariant with respect to transformation by years, from the theoretical point of view as well as in a sufficient statistic, or a diffeomorphic transformation of applications. In this approach, probability distributions the support in the nonparametric case (Cencov, 2000; are seen as elements of a differentiable manifold, on which Bauer et al., 2016). Arguably the most famous example a metric structure is defined through the choice of a of Fisher information geometry of a statistical model is Riemannian metric. Two very important ones are the that of the univariate Gaussian model, which is hyperbolic. Wasserstein metric, central in optimal transport, and The geometries of other parametric families such as the Fisher information metric (also called Fisher-Rao the multivariate Gaussian model (Atkinson and Mitchell, metric), essential in information geometry. Unlike optimal 1981; Skovgaard, 1984), the family of gamma distributions transport, information geometry is foremost concerned (Arwini and Dodson, 2008; Rebbah et al., 2019), or more with parametric families of probability distributions, and generally location-scale models (Said et al., 2019), among defines a Riemannian structure on the parameter space others, have also received a lot of attention.

We study the convergence rates of the EM algorithm for learning two-component mixed linear regression under all regimes of signal-to-noise ratio (SNR). We resolve a long-standing question that many recent results have attempted to tackle: we completely characterize the convergence behavior of EM, and show that the EM algorithm achieves minimax optimal sample complexity under all SNR regimes. In particular, when the SNR is sufficiently large, the EM updates converge to the true parameter $\theta^{*}$ at the standard parametric convergence rate $\mathcal{O}((d/n)^{1/2})$ after $\mathcal{O}(\log(n/d))$ iterations. In the regime where the SNR is above $\mathcal{O}((d/n)^{1/4})$ and below some constant, the EM iterates converge to a $\mathcal{O}({\rm SNR}^{-1} (d/n)^{1/2})$ neighborhood of the true parameter, when the number of iterations is of the order $\mathcal{O}({\rm SNR}^{-2} \log(n/d))$. In the low SNR regime where the SNR is below $\mathcal{O}((d/n)^{1/4})$, we show that EM converges to a $\mathcal{O}((d/n)^{1/4})$ neighborhood of the true parameters, after $\mathcal{O}((n/d)^{1/2})$ iterations. Notably, these results are achieved under mild conditions of either random initialization or an efficiently computable local initialization. By providing tight convergence guarantees of the EM algorithm in middle-to-low SNR regimes, we fill the remaining gap in the literature, and significantly, reveal that in low SNR, EM changes rate, matching the $n^{-1/4}$ rate of the MLE, a behavior that previous work had been unable to show.

We consider speeding up stochastic gradient descent (SGD) by parallelizing it across multiple workers. We assume the same data set is shared among $n$ workers, who can take SGD steps and coordinate with a central server. Unfortunately, this could require a lot of communication between the workers and the server, which can dramatically reduce the gains from parallelism. The Local SGD method, proposed and analyzed in the earlier literature, suggests machines should make many local steps between such communications. While the initial analysis of Local SGD showed it needs $\Omega ( \sqrt{T} )$ communications for $T$ local gradient steps in order for the error to scale proportionately to $1/(nT)$, this has been successively improved in a string of papers, with the state-of-the-art requiring $\Omega \left( n \left( \mbox{ polynomial in log } (T) \right) \right)$ communications. In this paper, we give a new analysis of Local SGD. A consequence of our analysis is that Local SGD can achieve an error that scales as $1/(nT)$ with only a fixed number of communications independent of $T$: specifically, only $\Omega(n)$ communications are required.

Drought is a natural creeping threat with numerous damaging effects in various aspects of human life. Accurate drought prediction is a promising step in helping policy makers to set drought risk management strategies. To fulfill this purpose, choosing appropriate models plays an important role in predicting approach. In this study, different models of Artificial Neural Network (ANN) are employed to predict short and long-term of droughts by using Standardized Precipitation Index (SPI) at different time scales, including 3, 6, 12, 24 and 48 months in Tabriz city, Iran. To this end, different combination of calculated SPI and time series of various hydro-meteorological variables, such as precipitation, wind velocity, relative humidity and sunshine hours for years 1992 to 2010 are used to train the ANN models. In order to compare the models performances, some well-known measures, namely RMSE, Mean Absolute Error (MAE) and Correlation Coefficient (CC) are utilized in the present study. The results illustrate that the application of all hydro-meteorological variables significantly improves the prediction of SPI at different time scales.

Conformal inference, cross-validation+, and the jackknife+ are hold-out methods that can be combined with virtually any machine learning algorithm to construct prediction sets with guaranteed marginal coverage. In this paper, we develop specialized versions of these techniques for categorical and unordered response labels that, in addition to providing marginal coverage, are also fully adaptive to complex data distributions, in the sense that they perform favorably in terms of approximate conditional coverage compared to alternative methods. The heart of our contribution is a novel conformity score, which we explicitly demonstrate to be powerful and intuitive for classification problems, but whose underlying principle is potentially far more general. Experiments on synthetic and real data demonstrate the practical value of our theoretical guarantees, as well as the statistical advantages of the proposed methods over the existing alternatives.