Results so far obtained allow us to predict the RainTomorrow Yes/No variable. As a consequence, we are able so far to predict if tomorrow rainfall shall be above 1mm or not. In case of "at least moderate" rainfall, we would like to be as much reliable as possible in predicting {RainTomorrow "Yes"}. Since RainTomorrow "Yes" is perceived as the prediction of a potential threat of damages due to the rainfall, we have to alert Canberra's citizens properly. That translates in having a very good specificity, as explained in the presecution of the analysis. That is motivated by the fact that weather forecast comprises more than one prediction.

What if our data have way more than 3-dimensions? In the table is the average consumption of 17 types of food in grams per person per week for every country in the UK. The table shows some interesting variations across different food types, but overall differences aren't so notable. Let's see if PCA can eliminate dimensions to emphasize how countries differ. Already we can see something is different about Northern Ireland.

Non-convex optimization with local search heuristics has been widely used in machine learning, achieving many state-of-art results. It becomes increasingly important to understand why they can work for these NP-hard problems on typical data. The landscape of many objective functions in learning has been conjectured to have the geometric property that "all local optima are (approximately) global optima", and thus they can be solved efficiently by local search algorithms. However, establishing such property can be very difficult. In this paper, we analyze the optimization landscape of the random over-complete tensor decomposition problem, which has many applications in unsupervised learning, especially in learning latent variable models. In practice, it can be efficiently solved by gradient ascent on a non-convex objective. We show that for any small constant $\epsilon > 0$, among the set of points with function values $(1+\epsilon)$-factor larger than the expectation of the function, all the local maxima are approximate global maxima. Previously, the best-known result only characterizes the geometry in small neighborhoods around the true components. Our result implies that even with an initialization that is barely better than the random guess, the gradient ascent algorithm is guaranteed to solve this problem. Our main technique uses Kac-Rice formula and random matrix theory. To our best knowledge, this is the first time when Kac-Rice formula is successfully applied to counting the number of local minima of a highly-structured random polynomial with dependent coefficients.

Rgtsvm provides a fast and flexible support vector machine (SVM) implementation for the R language. The distinguishing feature of Rgtsvm is that support vector classification and support vector regression tasks are implemented on a graphical processing unit (GPU), allowing the libraries to scale to millions of examples with >100-fold improvement in performance over existing implementations. Nevertheless, Rgtsvm retains feature parity and has an interface that is compatible with the popular e1071 SVM package in R. Altogether, Rgtsvm enables large SVM models to be created by both experienced and novice practitioners.

Quantum computing promises fast solutions to a wide range of optimization problems and thus holds considerable potential for machine learning [1]-[3]. However, while the quantum machine learning literature so far mainly focused on the quantum gate paradigm, noticeable technological progress leading to commercial devices is happening in adiabatic quantum computing [4], [5]. Current adiabatic quantum computers are geared towards solving quadratic unconstrained binary optimization problems or Ising models. A simple strategy for setting up established learning algorithms to run on such devices is therefore to attempt to (re-)formulate or approximate their minimization or maximization objectives in terms of Ising models. In this paper, we apply this strategy to a simple unsupervised learning problem, namely binary clustering.

In this paper we first identify a basic limitation in gradient descent-based optimization methods when used in conjunctions with smooth kernels. An analysis based on the spectral properties of the kernel demonstrates that only a vanishingly small portion of the function space is reachable after a polynomial number of gradient descent iterations. This lack of approximating power drastically limits gradient descent for a fixed computational budget leading to serious over-regularization/underfitting. The issue is purely algorithmic, persisting even in the limit of infinite data. To address this shortcoming in practice, we introduce EigenPro iteration, based on a preconditioning scheme using a small number of approximately computed eigenvectors. It can also be viewed as learning a new kernel optimized for gradient descent. It turns out that injecting this small (computationally inexpensive and SGD-compatible) amount of approximate second-order information leads to major improvements in convergence. For large data, this translates into significant performance boost over the standard kernel methods. In particular, we are able to consistently match or improve the state-of-the-art results recently reported in the literature with a small fraction of their computational budget. Finally, we feel that these results show a need for a broader computational perspective on modern large-scale learning to complement more traditional statistical and convergence analyses. In particular, many phenomena of large-scale high-dimensional inference are best understood in terms of optimization on infinite dimensional Hilbert spaces, where standard algorithms can sometimes have properties at odds with finite-dimensional intuition. A systematic analysis concentrating on the approximation power of such algorithms within a budget of computation may lead to progress both in theory and practice.

In this paper, we propose a general framework for tensor singular value decomposition (tensor SVD), which focuses on the methodology and theory for extracting the hidden low-rank structure from high-dimensional tensor data. Comprehensive results are developed on both the statistical and computational limits for tensor SVD. This problem exhibits three different phases according to the signal-noise-ratio (SNR). In particular, with strong SNR, we show that the classical higher order orthogonal iteration achieves the minimax optimal rate of convergence in estimation; with weak SNR, the information-theoretical lower bound implies that it is impossible to have consistent estimation in general; with moderate SNR, we show that the non-convex maximum likelihood estimation provides optimal solution, but with NP-hard computational cost; moreover, under the hardness hypothesis of hypergraphic planted clique detection, there are no polynomial-time algorithms performing consistently in general.

Random Projection (RP) technique has been widely applied in many scenarios because it can reduce high-dimensional features into low-dimensional space within short time and meet the need of real-time analysis of massive data. There is an urgent need of dimensionality reduction with fast increase of big genomics data. However, the performance of RP is usually lower. We attempt to improve classification accuracy of RP through combining other reduction dimension methods such as Principle Component Analysis (PCA), Linear Discriminant Analysis (LDA), and Feature Selection (FS). We compared classification accuracy and running time of different combination methods on three microarray datasets and a simulation dataset. Experimental results show a remarkable improvement of 14.77% in classification accuracy of FS followed by RP compared to RP on BC-TCGA dataset. LDA followed by RP also helps RP to yield a more discriminative subspace with an increase of 13.65% on classification accuracy on the same dataset. FS followed by RP outperforms other combination methods in classification accuracy on most of the datasets.

In this blog, we will discuss about clustering of customer activities for 24 hours by using K-means clustering feature in Tableau 10. This type of clustering helps you create statistically-based segments that provide insights about similarities in different groups and performance of the groups when compared to each other. You can use clustering on any type of visualization ranging from scatter plots to text tables and even maps. In our previous blog post – "Call Detail Record Analysis – K-means Clustering with R", we have discussed about CDR analysis using unsupervised K-means clustering algorithm. A daily activity file from Dandelion API is used as a data source, where the file contains CDR records generated by the Telecom Italia cellular network over the city of Milano.

The following simpler spectral partitioning approach (thresholding on the second dominant eigenvector) can also be applied for automatic separation of the foreground from the background.