In a world where nearly all manual tasks are being automated, the definition of manual is changing. Machine Learning algorithms can help computers play chess, perform surgeries, and get smarter and more personal. We are living in an era of constant technological progress, and looking at how computing has advanced over the years, we can predict what's to come in the days ahead. One of the main features of this revolution that stands out is how computing tools and techniques have been democratized. In the past five years, data scientists have built sophisticated data-crunching machines by seamlessly executing advanced techniques.

An Introduction to Statistical Learning, with Applications in R, written by Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani, is an absolute classic in the space. The book, a staple of statistical learning texts, is accessible to readers of all levels, and can be read without much of an existing foundational knowledge in the area. While the original has been around since 2013, the second edition was published very recently, and is now freely-available via PDF on the book's website. As the scale and scope of data collection continue to increase across virtually all fields, statistical learning has become a critical toolkit for anyone who wishes to understand data. An Introduction to Statistical Learning provides a broad and less technical treatment of key topics in statistical learning.

Kernel methods are fundamental in machine learning, and faster algorithms for kernel approximation provide direct speedups for many core tasks in machine learning. The polynomial kernel is especially important as other kernels can often be approximated by the polynomial kernel via a Taylor series expansion. Recent techniques in oblivious sketching reduce the dependence in the running time on the degree $q$ of the polynomial kernel from exponential to polynomial, which is useful for the Gaussian kernel, for which $q$ can be chosen to be polylogarithmic. However, for more slowly growing kernels, such as the neural tangent and arc-cosine kernels, $q$ needs to be polynomial, and previous work incurs a polynomial factor slowdown in the running time. We give a new oblivious sketch which greatly improves upon this running time, by removing the dependence on $q$ in the leading order term. Combined with a novel sampling scheme, we give the fastest algorithms for approximating a large family of slow-growing kernels.

Cardiovascular diseases and their associated disorder of heart failure are one of the major death causes globally, being a priority for doctors to detect and predict its onset and medical consequences. Artificial Intelligence (AI) allows doctors to discover clinical indicators and enhance their diagnosis and treatments. Specifically, explainable AI offers tools to improve the clinical prediction models that experience poor interpretability of their results. This work presents an explainability analysis and evaluation of a prediction model for heart failure survival by using a dataset that comprises 299 patients who suffered heart failure. The model employs a data workflow pipeline able to select the best ensemble tree algorithm as well as the best feature selection technique. Moreover, different post-hoc techniques have been used for the explainability analysis of the model. The paper's main contribution is an explainability-driven approach to select the best prediction model for HF survival based on an accuracy-explainability balance. Therefore, the most balanced explainable prediction model implements an Extra Trees classifier over 5 selected features (follow-up time, serum creatinine, ejection fraction, age and diabetes) out of 12, achieving a balanced-accuracy of 85.1% and 79.5% with cross-validation and new unseen data respectively. The follow-up time is the most influencing feature followed by serum-creatinine and ejection-fraction. The explainable prediction model for HF survival presented in this paper would improve a further adoption of clinical prediction models by providing doctors with intuitions to better understand the reasoning of, usually, black-box AI clinical solutions, and make more reasonable and data-driven decisions.

Consider the sequential optimization of a continuous, possibly non-convex, and expensive to evaluate objective function $f$. The problem can be cast as a Gaussian Process (GP) bandit where $f$ lives in a reproducing kernel Hilbert space (RKHS). The state of the art analysis of several learning algorithms shows a significant gap between the lower and upper bounds on the simple regret performance. When $N$ is the number of exploration trials and $\gamma_N$ is the maximal information gain, we prove an $\tilde{\mathcal{O}}(\sqrt{\gamma_N/N})$ bound on the simple regret performance of a pure exploration algorithm that is significantly tighter than the existing bounds. We show that this bound is order optimal up to logarithmic factors for the cases where a lower bound on regret is known. To establish these results, we prove novel and sharp confidence intervals for GP models applicable to RKHS elements which may be of broader interest.

This monograph develops a comprehensive statistical learning framework that is robust to (distributional) perturbations in the data using Distributionally Robust Optimization (DRO) under the Wasserstein metric. Beginning with fundamental properties of the Wasserstein metric and the DRO formulation, we explore duality to arrive at tractable formulations and develop finite-sample, as well as asymptotic, performance guarantees. We consider a series of learning problems, including (i) distributionally robust linear regression; (ii) distributionally robust regression with group structure in the predictors; (iii) distributionally robust multi-output regression and multiclass classification, (iv) optimal decision making that combines distributionally robust regression with nearest-neighbor estimation; (v) distributionally robust semi-supervised learning, and (vi) distributionally robust reinforcement learning. A tractable DRO relaxation for each problem is being derived, establishing a connection between robustness and regularization, and obtaining bounds on the prediction and estimation errors of the solution. Beyond theory, we include numerical experiments and case studies using synthetic and real data. The real data experiments are all associated with various health informatics problems, an application area which provided the initial impetus for this work.

Tóm tắt nội dung--A smart Ponzi scheme is a new form of economic crime that uses Ethereum smart contract account and cryptocurrency to implement Ponzi scheme. The smart Ponzi scheme has harmed the interests of many investors, but researches on smart Ponzi scheme detection is still very limited. The existing smart Ponzi scheme detection methods have the problems of requiring many human resources in feature engineering and poor model portability. To solve these problems, we propose a datadriven smart Ponzi scheme detection system in this paper. The system uses dynamic graph embedding technology to automatically learn the representation of an account based on multi-source and multi-modal data related to account transactions. Compared with traditional methods, the proposed system requires very limited human-computer interaction. To the best of our knowledge, this is the first work to implement smart Ponzi scheme detection through dynamic graph embedding. Ponzi schemes require a constant flow of funds from new investors. The detection method based on source contributed by new investors to pay off the returns of existing code inspection detects the smart Ponzi scheme by manually investors (Figure 1).

Federated learning aims to protect users' privacy while performing data analysis from different participants. However, it is challenging to guarantee the training efficiency on heterogeneous systems due to the various computational capabilities and communication bottlenecks. In this work, we propose FedSkel to enable computation-efficient and communication-efficient federated learning on edge devices by only updating the model's essential parts, named skeleton networks. FedSkel is evaluated on real edge devices with imbalanced datasets. Experimental results show that it could achieve up to 5.52$\times$ speedups for CONV layers' back-propagation, 1.82$\times$ speedups for the whole training process, and reduce 64.8% communication cost, with negligible accuracy loss.

IN FEBRUARY, Jaguar Land Rover--based in Gaydon, Warwickshire, and owned by the Tata Group of Mumbai--announced that the Leaping Cat was to be "reimagined" as an "all-electric luxury brand from 2025." In the months since, JLR has declined to provide details. But if realized in full, such a plan would carry Jaguar over the all-electric threshold years earlier than rivals including BMW, Mercedes-Benz and Audi. Uniquely, JLR's announcement seems to imply some sort of hard stop on petrol-powered Jaguars (about 426,000 global sales in 2020). Other automakers--Daimler and Toyota among them--want to ease the pain of electrification with a generation of gas-electric hybrids and plug-in hybrid models.

The vast majority of data science and machine learning models rely on creating a vector, or embedding, of your data. Some of these embeddings naturally create themselves. For example, for numerical data organized in columns we can think of the values associated with each row as a single vector. In more complicated cases such as natural language processing we have to generate those embeddings from the words through a variety of different approaches like one-hot encoding, skip-gram methods such as word2vec, etc. These vectors are then used as the representation of the data that is to be modeled.