Feature selection is a powerful dimension reduction technique which selects a subset of relevant features for model construction. Numerous feature selection methods have been proposed, but most of them fail under the high-dimensional and low-sample size (HDLSS) setting due to the challenge of overfitting. In this paper, we present a deep learning-based method - GRAph Convolutional nEtwork feature Selector (GRACES) - to select important features for HDLSS data. We demonstrate empirical evidence that GRACES outperforms other feature selection methods on both synthetic and real-world datasets.

While the comparison table in this article applies to a specific problem in FinTech, the conclusions are consistent with findings in other frameworks. There is no single method that outperforms all the other ones, for obvious reasons. To be the global winner means winning on all potential datasets. The immense majority of datasets are pure noise, for which performance is meaningless, and the winner randomly changes from one dataset to another. Yet, in real-life applications, each dataset comes with patterns or meaningful signal.

With everything going computerized and digital, the amount of data generated by us is humongous. Organizations collectively spend billions of dollars to just store and analyze this data. They make efforts to drive valuable business insights from this data using data mining. Data Mining is the process of discovering hidden patterns in a pile of big data. Business executives use these emerging patterns to make informed business strategy decisions.

In this blog, we have gathered the top 10 machine learning books. Learning this subject is a challenge for beginners. Take your learning experience one step ahead with these top-rated ML books on Amazon. Machine Learning: 4 Books in 1 is a complete guide for beginners to master the basics of Python programming and understand how to build artificial intelligence through data science. This book includes four books: Introduction to Machine Learning, Python Programming for Beginners, Data Science for Beginners, and Artificial Intelligence for Beginners.

Machine Learning for Probabilistic Prediction Quantitative Finance Webinar, Stony Brook University (11/11/2022) Valery Manokhin, PhD, MBA, CFQ Speaker Bio • PhD in Machine Learning (2022) from Royal Holloway, University of London • During PhD conducted research and published papers in probabilistic and conformal prediction. PhD supervised by Prof. Vladimir Vovk, the creator of Conformal Prediction (Prof. Vladimir Vovk is the last PhD student of Andrey Kolmogorov) • Dr. Valery Manokhin holds a number of advanced MSc degrees including from the Moscow Institute of Physics and Technology (Physics/Math), UCL (Computational Statistics and Machine Learning), University of Sussex (Quant Finance) and an MBA from the University of Warwick • Published in the leading machine learning journals, including'Neurocomputing', 'Journal of Machine Learning Research' and'Machine Learning Journal', also in the industry journals including'Frontiers in Energy Research' • Created'Awesome Conformal Prediction' - the most comprehensive professionally curated resource on Conformal Prediction (over 900 stars on GitHub). 'Awesome Conformal Prediction' has been featured at the leading conferences such as ICML and in Kevin Murphy's bestselling book'Probabilistic Machine Learning: An Introduction' Outline of this webinar Introduction to Probabilistic Prediction Probability Calibration Introduction to Conformal Prediction Conformal Prediction for Classification Conformal Prediction for Regression Conclusion 3 Why Probabilistic Prediction? Machine Learning is primarily concerned with producing functions mapping objects onto predicted labels Classical statistical techniques - for small scale, low-dimensional data High-dimensional data does not necessarily follow well-known distributions and hence required new approaches (e.g.

Machine learning (ML) models, and deep neural networks (DNNs) in particular, are very successful at solving problems both within and outside of science; the latest, spectacular scientific example is the prediction of protein folding (Jumper et al., 2021). However, many of these models are black boxes, and we do not know why they are so successful. As a consequence, the interpretability of ML models - understanding or gaining insight into how they work - is an important area of research in computer science. One kind of effort is towards a better grasp of theoretical properties of ML models, and to formulate what is called a theory of deep learning (Berner et al., 2021; Bahri et al., 2020). Another kind of effort is to provide ML practitioners with tools to understand predictions made by the ML models they deploy. This latter effort often runs under the label of explainable AI (xAI, see, e.g., Adadi and Berrada 2018). Philosophers have also started to pay more attention to interpretability recently; see Beisbart and Räz (2022) for a survey.

In this paper, we move towards combining large parametric models with non-parametric prototypical networks. We propose prototypical fine-tuning, a novel prototypical framework for fine-tuning pretrained language models (LM), which automatically learns a bias to improve predictive performance for varying data sizes, especially low-resource settings. Our prototypical fine-tuning approach can automatically adjust the model capacity according to the number of data points and the model's inherent attributes. Moreover, we propose four principles for effective prototype fine-tuning towards the optimal solution. Experimental results across various datasets show that our work achieves significant performance improvements under various low-resource settings, as well as comparable and usually better performances in high-resource scenarios.

Multi-Task Learning (MTL) is a powerful learning paradigm to improve generalization performance via knowledge sharing. However, existing studies find that MTL could sometimes hurt generalization, especially when two tasks are less correlated. One possible reason that hurts generalization is spurious correlation, i.e., some knowledge is spurious and not causally related to task labels, but the model could mistakenly utilize them and thus fail when such correlation changes. In MTL setup, there exist several unique challenges of spurious correlation. First, the risk of having non-causal knowledge is higher, as the shared MTL model needs to encode all knowledge from different tasks, and causal knowledge for one task could be potentially spurious to the other. Second, the confounder between task labels brings in a different type of spurious correlation to MTL. We theoretically prove that MTL is more prone to taking non-causal knowledge from other tasks than single-task learning, and thus generalize worse. To solve this problem, we propose Multi-Task Causal Representation Learning framework, aiming to represent multi-task knowledge via disentangled neural modules, and learn which module is causally related to each task via MTL-specific invariant regularization. Experiments show that it could enhance MTL model's performance by 5.5% on average over Multi-MNIST, MovieLens, Taskonomy, CityScape, and NYUv2, via alleviating spurious correlation problem.

In this paper we study the practicality and usefulness of incorporating distributed representations of graphs into models within the context of drug pair scoring. We argue that the real world growth and update cycles of drug pair scoring datasets subvert the limitations of transductive learning associated with distributed representations. Furthermore, we argue that the vocabulary of discrete substructure patterns induced over drug sets is not dramatically large due to the limited set of atom types and constraints on bonding patterns enforced by chemistry. Under this pretext, we explore the effectiveness of distributed representations of the molecular graphs of drugs in drug pair scoring tasks such as drug synergy, polypharmacy, and drug-drug interaction prediction. To achieve this, we present a methodology for learning and incorporating distributed representations of graphs within a unified framework for drug pair scoring. Subsequently, we augment a number of recent and state-of-the-art models to utilise our embeddings. We empirically show that the incorporation of these embeddings improves downstream performance of almost every model across different drug pair scoring tasks, even those the original model was not designed for. We publicly release all of our drug embeddings for the DrugCombDB, DrugComb, DrugbankDDI, and TwoSides datasets.

No human drives a car in a vacuum; she/he must negotiate with other road users to achieve their goals in social traffic scenes. A rational human driver can interact with other road users in a socially-compatible way through implicit communications to complete their driving tasks smoothly in interaction-intensive, safety-critical environments. This paper aims to review the existing approaches and theories to help understand and rethink the interactions among human drivers toward social autonomous driving. We take this survey to seek the answers to a series of fundamental questions: 1) What is social interaction in road traffic scenes? 2) How to measure and evaluate social interaction? 3) How to model and reveal the process of social interaction? 4) How do human drivers reach an implicit agreement and negotiate smoothly in social interaction? This paper reviews various approaches to modeling and learning the social interactions between human drivers, ranging from optimization theory and graphical models to social force theory and behavioral & cognitive science. We also highlight some new directions, critical challenges, and opening questions for future research.