sktime is an open source, Python based, sklearn compatible toolkit for time series analysis developed by researchers at the University of East Anglia, University College London and the Alan Turing Institute. A key initial goal for sktime was to provide time series classification functionality equivalent to that available in a related java package, tsml. We describe the implementation of six such classifiers in sktime and compare them to their tsml equivalents. We demonstrate correctness through equivalence of accuracy on a range of standard test problems and compare the build time of the different implementations. We find that there is significant difference in accuracy on only one of the six algorithms, and this difference was to be expected. We found a much wider range of difference in efficiency. Again, this was not unexpected, but it does highlight ways both toolkits could be improved. PLEASE NOTE THIS PAPER IS NOT COMPLETE. It is a work in progress and we have pushed it early so that we can reference it in another paper. More to follow!

Emergence of artificial intelligence techniques in biomedical applications urges the researchers to pay more attention on the uncertainty quantification (UQ) in machine-assisted medical decision making. For classification tasks, prior studies on UQ are difficult to compare with each other, due to the lack of a unified quantitative evaluation metric. Considering that well-performing UQ models ought to know when the classification models act incorrectly, we design a new evaluation metric, area under Confidence-Classification Characteristic curves (AUCCC), to quantitatively evaluate the performance of the UQ models. AUCCC is threshold-free, robust to perturbation, and insensitive to the classification performance. We evaluate several UQ methods (e.g., max softmax output) with AUCCC to validate its effectiveness. Furthermore, a simple scheme, named Uncertainty Distillation (UDist), is developed to boost the UQ performance, where a confidence model is distilling the confidence estimated by deep ensembles. The proposed method is easy to implement; it consistently outperforms strong baselines on natural and medical image datasets in our experiments.

This simple introduction to matrix theory offers a refreshing perspective on the subject. Using a basic concept that leads to a simple formula for the power of a matrix, we see how it can solve time series, Markov chains, linear regression, data reduction, principal components analysis (PCA) and other machine learning problems. These problems are usually solved with more advanced matrix calculus, including eigenvalues, diagonalization, generalized inverse matrices, and other types of matrix normalization. Our approach is more intuitive and thus appealing to professionals who do not have a strong mathematical background, or who have forgotten what they learned in math textbooks. It will also appeal to physicists and engineers.

As humans we are able to discern differences among different groups within a collection. We might group a collection by broad groups such as birds versus plants versus animals or detect subtle features to identify different makes and models of cars. Clustering techniques allow us to automate the process and apply them to data where groupings are not immediately obvious. These techniques are used for different purposes such as detecting market segments, identifying properties of online communities, fraud detection, and cybersecurity. Srini Anand is a Data Scientist at Ameritas Life Insurance Company and holds a Masters degree in Data Science from Indiana University.

Lecture notes for the Statistical Machine Learning course taught at the Department of Information Technology, University of Uppsala (Sweden.) Available as a PDF, here (original) or here (mirror). B.1 A general iterative solution B.2 Commonly used search directions

In many scientific and engineering applications, we are tasked with the maximisation of an expensive to evaluate black box function f. Traditional settings for this problem assume just the availability of this single function. However, in many cases, cheap approximations to f may be obtainable. For example, the expensive real world behaviour of a robot can be approximated by a cheap computer simulation. We can use these approximations to eliminate low function value regions cheaply and use the expensive evaluations of f in a small but promising region and speedily identify the optimum. We formalise this task as a multi-fidelity bandit problem where the target function and its approximations are sampled from a Gaussian process. We develop MF-GP-UCB, a novel method based on upper confidence bound techniques. In our theoretical analysis we demonstrate that it exhibits precisely the above behaviour and achieves better bounds on the regret than strategies which ignore multi-fidelity information. Empirically, MF-GP-UCB outperforms such naive strategies and other multi-fidelity methods on several synthetic and real experiments.

We study the problem of online clustering where a clustering algorithm has to assign a new point that arrives to one of $k$ clusters. The specific formulation we use is the $k$-means objective: At each time step the algorithm has to maintain a set of k candidate centers and the loss incurred is the squared distance between the new point and the closest center. The goal is to minimize regret with respect to the best solution to the $k$-means objective ($\mathcal{C}$) in hindsight. We show that provided the data lies in a bounded region, an implementation of the Multiplicative Weights Update Algorithm (MWUA) using a discretized grid achieves a regret bound of $\tilde{O}(\sqrt{T})$ in expectation. We also present an online-to-offline reduction that shows that an efficient no-regret online algorithm (despite being allowed to choose a different set of candidate centres at each round) implies an offline efficient algorithm for the $k$-means problem. In light of this hardness, we consider the slightly weaker requirement of comparing regret with respect to $(1 + \epsilon) \mathcal{C}$ and present a no-regret algorithm with runtime $O\left(T(\mathrm{poly}(log(T),k,d,1/\epsilon)^{k(d+O(1))}\right)$. Our algorithm is based on maintaining an incremental coreset and an adaptive variant of the MWUA. We show that na\"{i}ve online algorithms, such as \emph{Follow The Leader}, fail to produce sublinear regret in the worst case. We also report preliminary experiments with synthetic and real-world data.

In semi-supervised learning, virtual adversarial training (VAT) approach is one of the most attractive method due to its intuitional simplicity and powerful performances. VAT finds a classifier which is robust to data perturbation toward the adversarial direction. In this study, we provide a fundamental explanation why VAT works well in semi-supervised learning case and propose new techniques which are simple but powerful to improve the VAT method. Especially we employ the idea of Bad GAN approach, which utilizes bad samples distributed on complement of the support of the input data, without any additional deep generative architectures. We generate bad samples of high-quality by use of the adversarial training used in VAT and also give theoretical explanations why the adversarial training is good at both generating bad samples. An advantage of our proposed method is to achieve the competitive performances compared with other recent studies with much fewer computations. We demonstrate advantages our method by various experiments with well known benchmark image datasets.

For example, a person can be uniquely identified in terms of face, fingerprint, iris, and signature; an image can be described by different kinds of descriptors: SIFT, HOG, and LBP, where SIFT is robust to image illumination, noise, and rotation, HOG is sensitive to marginal information, while LBP is a powerful texture feature; the same document can be represented in different languages. Different views can capture distinct perspectives of data. Numerous real-world applications have benefited from multi-view data by leveraging the complementary information [5, 6, 7, 8, 9]. Thus, multi-view learning has become an important research field [10, 11]. As an important ingredient of multi-view learning, multi-view clustering has been widely investigated to identify underlying structures in multi-view data in an unsupervised way [12, 13]. Although each view contains different fractional information, they together admit the same clustering structure. Simply concatenating all features into a single view and then employing a clustering algorithm on this single view data might not obtain better performance than traditional methods which use single view separately [14, 11]. In the past decade, plenty of advanced multi-view clustering algorithms have been proposed and they perform effectively by considering the diversity and complementarity of different views.

In this paper, we focus on the separability of classes with the cross-entropy loss function for classification problems by theoretically analyzing the intra-class distance and inter-class distance (i.e. the distance between any two points belonging to the same class and different classes, respectively) in the feature space, i.e. the space of representations learnt by neural networks. Specifically, we consider an arbitrary network architecture having a fully connected final layer with Softmax activation and trained using the cross-entropy loss. We derive expressions for the value and the distribution of the squared L2 norm of the product of a network dependent matrix and a random intra-class and inter-class distance vector (i.e. the vector between any two points belonging to the same class and different classes), respectively, in the learnt feature space (or the transformation of the original data) just before Softmax activation, as a function of the cross-entropy loss value. The main result of our analysis is the derivation of a lower bound for the probability with which the inter-class distance is more than the intra-class distance in this feature space, as a function of the loss value. We do so by leveraging some empirical statistical observations with mild assumptions and sound theoretical analysis. As per intuition, the probability with which the inter-class distance is more than the intra-class distance decreases as the loss value increases, i.e. the classes are better separated when the loss value is low. To the best of our knowledge, this is the first work of theoretical nature trying to explain the separability of classes in the feature space learnt by neural networks trained with the cross-entropy loss function.