Machine Learning最新文献

Markov chains play a key role in a vast number of areas, including life insurance mathematics. Standard actuarial quantities as the premium value can be interpreted as compressed, lossy information about the underlying Markov process. We introduce a method to reconstruct the underlying Markov chain given collective information of a portfolio of contracts. Our neural architecture characterizes the process in a highly explainable way by explicitly providing one-step transition probabilities. Further, we provide an intrinsic, economic model validation to inspect the quality of the information decompression. Lastly, our methodology is successfully tested for a realistic data set of German term life insurance contracts.

Aging involves complex biological processes leading to the decline of living organisms. As population lifespan increases worldwide, the importance of identifying factors underlying healthy aging has become critical. Integration of multi-modal datasets is a powerful approach for the analysis of complex biological systems, with the potential to uncover novel aging biomarkers. In this study, we leveraged publicly available epigenomic, transcriptomic and telomere length data along with histological images from the Genotype-Tissue Expression project to build tissue-specific regression models for age prediction. Using data from two tissues, lung and ovary, we aimed to compare model performance across data modalities, as well as to assess the improvement resulting from integrating multiple data types. Our results demostrate that methylation outperformed the other data modalities, with a mean absolute error of 3.36 and 4.36 in the test sets for lung and ovary, respectively. These models achieved lower error rates when compared with established state-of-the-art tissue-agnostic methylation models, emphasizing the importance of a tissue-specific approach. Additionally, this work has shown how the application of Hierarchical Image Pyramid Transformers for feature extraction significantly enhances age modeling using histological images. Finally, we evaluated the benefits of integrating multiple data modalities into a single model. Combining methylation data with other data modalities only marginally improved performance likely due to the limited number of available samples. Combining gene expression with histological features yielded more accurate age predictions compared with the individual performance of these data types. Given these results, this study shows how machine learning applications can be extended to/in multi-modal aging research. Code used is available at https://github.com/zroger49/multi_modal_age_prediction.

Finding dense subgraphs is a core problem in graph mining with many applications in diverse domains. At the same time many real-world networks vary over time, that is, the dataset can be represented as a sequence of graph snapshots. Hence, it is natural to consider the question of finding dense subgraphs in a temporal network that are allowed to vary over time to a certain degree. In this paper, we search for dense subgraphs that have large pairwise Jaccard similarity coefficients. More formally, given a set of graph snapshots and input parameter (alpha), we find a collection of dense subgraphs, with pairwise Jaccard index at least (alpha), such that the sum of densities of the induced subgraphs is maximized. We prove that this problem is NP-hard and we present a greedy, iterative algorithm which runs in ({mathcal {O}} mathopen {} left( nk^2 + mright)) time per single iteration, where k is the length of the graph sequence and n and m denote number of vertices and total number of edges respectively. We also consider an alternative problem where subgraphs with large pairwise Jaccard indices are rewarded. We do this by incorporating the indices directly into the objective function. More formally, given a set of graph snapshots and a weight (lambda), we find a collection of dense subgraphs such that the sum of densities of the induced subgraphs plus the sum of Jaccard indices, weighted by (lambda), is maximized. We prove that this problem is NP-hard. To discover dense subgraphs with good objective value, we present an iterative algorithm which runs in ({mathcal {O}} mathopen {}left( n^2k^2 + m log n + k^3 nright)) time per single iteration, and a greedy algorithm which runs in ({mathcal {O}} mathopen {}left( n^2k^2 + m log n + k^3 nright)) time. We show experimentally that our algorithms are efficient, they can find ground truth in synthetic datasets and provide good results from real-world datasets. Finally, we present two case studies that show the usefulness of our problem.

Sequences of labeled events observed at irregular intervals in continuous time are ubiquitous across various fields. Temporal Point Processes (TPPs) provide a mathematical framework for modeling these sequences, enabling inferences such as predicting the arrival time of future events and their associated label, called mark. However, due to model misspecification or lack of training data, these probabilistic models may provide a poor approximation of the true, unknown underlying process, with prediction regions extracted from them being unreliable estimates of the underlying uncertainty. This paper develops more reliable methods for uncertainty quantification in neural TPP models via the framework of conformal prediction. A primary objective is to generate a distribution-free joint prediction region for an event’s arrival time and mark, with a finite-sample marginal coverage guarantee. A key challenge is to handle both a strictly positive, continuous response and a categorical response, without distributional assumptions. We first consider a simple but overly conservative approach that combines individual prediction regions for the event’s arrival time and mark. Then, we introduce a more effective method based on bivariate highest density regions derived from the joint predictive density of arrival times and marks. By leveraging the dependencies between these two variables, this method excludes unlikely combinations of the two, resulting in sharper prediction regions while still attaining the pre-specified coverage level. We also explore the generation of individual univariate prediction regions for events’ arrival times and marks through conformal regression and classification techniques. Moreover, we evaluate the stronger notion of conditional coverage. Finally, through extensive experimentation on both simulated and real-world datasets, we assess the validity and efficiency of these methods.

We propose a new machine learning formulation designed specifically for extrapolation. The textbook way to apply machine learning to drug design is to learn a univariate function that when a drug (structure) is input, the function outputs a real number (the activity): f(drug) (rightarrow) activity. However, experience in real-world drug design suggests that this formulation of the drug design problem is not quite correct. Specifically, what one is really interested in is extrapolation: predicting the activity of new drugs with higher activity than any existing ones. Our new formulation for extrapolation is based on learning a bivariate function that predicts the difference in activities of two drugs F(drug1, drug2) (rightarrow) difference in activity, followed by the use of ranking algorithms. This formulation is general and agnostic, suitable for finding samples with target values beyond the target value range of the training set. We applied the formulation to work with support vector machines , random forests , and Gradient Boosting Machines . We compared the formulation with standard regression on thousands of drug design datasets, gene expression datasets and material property datasets. The test set extrapolation metric was the identification of examples with greater values than the training set, and top-performing examples (within the top 10% of the whole dataset). On this metric our pairwise formulation vastly outperformed standard regression. Its proposed variations also showed a consistent outperformance. Its application in the stock selection problem further confirmed the advantage of this pairwise formulation.

This paper introduces an Automated Machine Learning (AutoML) framework specifically designed to efficiently synthesize end-to-end multimodal machine learning pipelines. Traditional reliance on the computationally demanding Neural Architecture Search is minimized through the strategic integration of pre-trained transformer models. This innovative approach enables the effective unification of diverse data modalities into high-dimensional embeddings, streamlining the pipeline development process. We leverage an advanced Bayesian Optimization strategy, informed by meta-learning, to facilitate the warm-starting of the pipeline synthesis, thereby enhancing computational efficiency. Our methodology demonstrates its potential to create advanced and custom multimodal pipelines within limited computational resources. Extensive testing across 23 varied multimodal datasets indicates the promise and utility of our framework in diverse scenarios. The results contribute to the ongoing efforts in the AutoML field, suggesting new possibilities for efficiently handling complex multimodal data. This research represents a step towards developing more efficient and versatile tools in multimodal machine learning pipeline development, acknowledging the collaborative and ever-evolving nature of this field.

This study builds upon our previous work by introducing a refined Inductive Conformal Martingale (ICM) approach for addressing Concept Drift. Specifically, we enhance our previously proposed CAUTIOUS betting function to incorporate multiple density estimators for improving detection ability. We also combine this betting function with two base estimators that have not been previously utilized within the ICM framework: the Interpolated Histogram and Nearest Neighbor Density Estimators. We assess these extensions using both a single ICM and an ensemble of ICMs. For the latter, we conduct a comprehensive experimental investigation into the influence of the ensemble size on prediction accuracy and the number of available predictions. Our experimental results on four benchmark datasets demonstrate that the proposed approach surpasses our previous methodology in terms of performance while matching or in many cases exceeding that of three contemporary state-of-the-art techniques.

Datasets comprised of large sets of both predictor and outcome variables are becoming more widely used in research. In addition to the well-known problems of model complexity and predictor variable selection, predictive modelling with such large data also presents a relatively novel and under-studied challenge of outcome variable selection. Certain outcome variables in the data may not be adequately predicted by the given sets of predictors. In this paper, we propose the method of Sparse Multivariate Principal Covariates Regression that addresses these issues altogether by expanding the Principal Covariates Regression model to incorporate sparsity penalties on both of predictor and outcome variables. Our method is one of the first methods that perform variable selection for both predictors and outcomes simultaneously. Moreover, by relying on summary variables that explain the variance in both predictor and outcome variables, the method offers a sparse and succinct model representation of the data. In a simulation study, the method performed better than methods with similar aims such as sparse Partial Least Squares at prediction of the outcome variables and recovery of the population parameters. Lastly, we administered the method on an empirical dataset to illustrate its application in practice.

There has been an explosion of data collected about sports. Because such data is extremely rich and complex, machine learning is increasingly being used to extract actionable insights from it. Typically, machine learning is used to build models and indicators that capture the skills, capabilities, and tendencies of athletes and teams. Such indicators and models are in turn used to inform decision-making at professional clubs. Designing these indicators requires paying careful attention to a number of subtle issues from a methodological and evaluation perspective. In this paper, we highlight these challenges in sports and discuss a variety of approaches for handling them. Methodologically, we highlight that dependencies affect how to perform data partitioning for evaluation as well as the need to consider contextual factors. From an evaluation perspective, we draw a distinction between evaluating the developed indicators themselves versus the underlying models that power them. We argue that both aspects must be considered, but that they require different approaches. We hope that this article helps bridge the gap between traditional sports expertise and modern data analytics by providing a structured framework with practical examples.

Recent work on Vision Transformers (VTs) showed that introducing a local inductive bias in the VT architecture helps reducing the number of samples necessary for training. However, the architecture modifications lead to a loss of generality of the Transformer backbone, partially contradicting the push towards the development of uniform architectures, shared, e.g., by both the Computer Vision and the Natural Language Processing areas. In this work, we propose a different and complementary direction, in which a local bias is introduced using an auxiliary self-supervised task, performed jointly with standard supervised training. Specifically, we exploit the observation that the attention maps of VTs, when trained with self-supervision, can contain a semantic segmentation structure which does not spontaneously emerge when training is supervised. Thus, we explicitly encourage the emergence of this spatial clustering as a form of training regularization. In more detail, we exploit the assumption that, in a given image, objects usually correspond to few connected regions, and we propose a spatial formulation of the information entropy to quantify this object-based inductive bias. By minimizing the proposed spatial entropy, we include an additional self-supervised signal during training. Using extensive experiments, we show that the proposed regularization leads to equivalent or better results than other VT proposals which include a local bias by changing the basic Transformer architecture, and it can drastically boost the VT final accuracy when using small-medium training sets. The code is available at https://github.com/helia95/SAR.