Semi-structured regression models enable the joint modeling of interpretable structured and complex unstructured feature effects. The structured model part is inspired by statistical models and can be used to infer the input-output relationship for features of particular importance. The complex unstructured part defines an arbitrary deep neural network and thereby provides enough flexibility to achieve competitive prediction performance. While these models can also account for aleatoric uncertainty, there is still a lack of work on accounting for epistemic uncertainty. In this paper, we address this problem by presenting a Bayesian approximation for semi-structured regression models using subspace inference. To this end, we extend subspace inference for joint posterior sampling from a full parameter space for structured effects and a subspace for unstructured effects. Apart from this hybrid sampling scheme, our method allows for tunable complexity of the subspace and can capture multiple minima in the loss landscape. Numerical experiments validate our approach's efficacy in recovering structured effect parameter posteriors in semi-structured models and approaching the full-space posterior distribution of MCMC for increasing subspace dimension. Further, our approach exhibits competitive predictive performance across simulated and real-world datasets.

This paper studies offline reinforcement learning with linear function approximation in a setting with decision-theoretic, but not estimation sparsity. The structural restrictions of the data-generating process presume that the transitions factor into a sparse component that affects the reward and could affect additional exogenous dynamics that do not affect the reward. Although the minimally sufficient adjustment set for estimation of full-state transition properties depends on the whole state, the optimal policy and therefore state-action value function depends only on the sparse component: we call this causal/decision-theoretic sparsity. We develop a method for reward-filtering the estimation of the state-action value function to the sparse component by a modification of thresholded lasso in least-squares policy evaluation. We provide theoretical guarantees for our reward-filtered linear fitted-Q-iteration, with sample complexity depending only on the size of the sparse component.

When selecting data for training large-scale models, standard practice is to filter for examples that match human notions of data quality. Such filtering yields qualitatively clean datapoints that intuitively should improve model behavior. However, in practice the opposite can often happen: we find that selecting according to similarity with "high quality" data sources may not increase (and can even hurt) performance compared to randomly selecting data. To develop better methods for selecting data, we start by framing dataset selection as an optimization problem that we can directly solve for: given target tasks, a learning algorithm, and candidate data, select the subset that maximizes model performance. This framework thus avoids handpicked notions of data quality, and instead models explicitly how the learning process uses train datapoints to predict on the target tasks. Our resulting method greatly improves language model (LM) performance on both pre-specified tasks and previously unseen tasks. Specifically, choosing target tasks representative of standard LM problems and evaluating on diverse held-out benchmarks, our selected datasets provide a 2x compute multiplier over baseline methods.

This article delves into the analysis of performance and utilization of Support Vector Machines (SVMs) for the critical task of forest fire detection using image datasets. With the increasing threat of forest fires to ecosystems and human settlements, the need for rapid and accurate detection systems is of utmost importance. SVMs, renowned for their strong classification capabilities, exhibit proficiency in recognizing patterns associated with fire within images. By training on labeled data, SVMs acquire the ability to identify distinctive attributes associated with fire, such as flames, smoke, or alterations in the visual characteristics of the forest area. The document thoroughly examines the use of SVMs, covering crucial elements like data preprocessing, feature extraction, and model training. It rigorously evaluates parameters such as accuracy, efficiency, and practical applicability. The knowledge gained from this study aids in the development of efficient forest fire detection systems, enabling prompt responses and improving disaster management. Moreover, the correlation between SVM accuracy and the difficulties presented by high-dimensional datasets is carefully investigated, demonstrated through a revealing case study. The relationship between accuracy scores and the different resolutions used for resizing the training datasets has also been discussed in this article. These comprehensive studies result in a definitive overview of the difficulties faced and the potential sectors requiring further improvement and focus.

The classification of different grapevine varieties is a relevant phenotyping task in Precision Viticulture since it enables estimating the growth of vineyard rows dedicated to different varieties, among other applications concerning the wine industry. This task can be performed with destructive methods that require time-consuming tasks, including data collection and analysis in the laboratory. However, Unmanned Aerial Vehicles (UAV) provide a more efficient and less prohibitive approach to collecting hyperspectral data, despite acquiring noisier data. Therefore, the first task is the processing of these data to correct and downsample large amounts of data. In addition, the hyperspectral signatures of grape varieties are very similar. In this work, a Convolutional Neural Network (CNN) is proposed for classifying seventeen varieties of red and white grape variants. Rather than classifying single samples, these are processed together with their neighbourhood. Hence, the extraction of spatial and spectral features is addressed with 1) a spatial attention layer and 2) Inception blocks. The pipeline goes from processing to dataset elaboration, finishing with the training phase. The fitted model is evaluated in terms of response time, accuracy and data separability, and compared with other state-of-the-art CNNs for classifying hyperspectral data. Our network was proven to be much more lightweight with a reduced number of input bands, a lower number of trainable weights and therefore, reduced training time. Despite this, the evaluated metrics showed much better results for our network (~99% overall accuracy), in comparison with previous works barely achieving 81% OA.

Speaker diarization, the task of segmenting an audio recording based on speaker identity, constitutes an important speech pre-processing step for several downstream applications. The conventional approach to diarization involves multiple steps of embedding extraction and clustering, which are often optimized in an isolated fashion. While end-to-end diarization systems attempt to learn a single model for the task, they are often cumbersome to train and require large supervised datasets. In this paper, we propose an end-to-end supervised hierarchical clustering algorithm based on graph neural networks (GNN), called End-to-end Supervised HierARchical Clustering (E-SHARC). The E-SHARC approach uses front-end mel-filterbank features as input and jointly learns an embedding extractor and the GNN clustering module, performing representation learning, metric learning, and clustering with end-to-end optimization. Further, with additional inputs from an external overlap detector, the E-SHARC approach is capable of predicting the speakers in the overlapping speech regions. The experimental evaluation on several benchmark datasets like AMI, VoxConverse and DISPLACE, illustrates that the proposed E-SHARC framework improves significantly over the state-of-art diarization systems.

We introduce and investigate the iterated application of Generalized Matrix Learning Vector Quantizaton for the analysis of feature relevances in classification problems, as well as for the construction of class-discriminative subspaces. The suggested Iterated Relevance Matrix Analysis (IRMA) identifies a linear subspace representing the classification specific information of the considered data sets using Generalized Matrix Learning Vector Quantization (GMLVQ). By iteratively determining a new discriminative subspace while projecting out all previously identified ones, a combined subspace carrying all class-specific information can be found. This facilitates a detailed analysis of feature relevances, and enables improved low-dimensional representations and visualizations of labeled data sets. Additionally, the IRMA-based classdiscriminative subspace can be used for dimensionality reduction and the training of robust classifiers with potentially improved performance. Preprint submitted to Neurocomputing January 24, 2024 1. Introduction Prototype-based systems such as Learning Vector Quantization (LVQ) [1, 2, 3, 4] can serve as genuinely interpretable and transparent classification tools [5].

Successive proposals of several self-supervised training schemes continue to emerge, taking one step closer to developing a universal foundation model. In this process, the unsupervised downstream tasks are recognized as one of the evaluation methods to validate the quality of visual features learned with a self-supervised training scheme. However, unsupervised dense semantic segmentation has not been explored as a downstream task, which can utilize and evaluate the quality of semantic information introduced in patch-level feature representations during self-supervised training of a vision transformer. Therefore, this paper proposes a novel data-driven approach for unsupervised semantic segmentation (DatUS^2) as a downstream task. DatUS^2 generates semantically consistent and dense pseudo annotate segmentation masks for the unlabeled image dataset without using any visual-prior or synchronized data. We compare these pseudo-annotated segmentation masks with ground truth masks for evaluating recent self-supervised training schemes to learn shared semantic properties at the patch level and discriminative semantic properties at the segment level. Finally, we evaluate existing state-of-the-art self-supervised training schemes with our proposed downstream task, i.e., DatUS^2. Also, the best version of DatUS^2 outperforms the existing state-of-the-art method for the unsupervised dense semantic segmentation task with 15.02% MiOU and 21.47% Pixel accuracy on the SUIM dataset. It also achieves a competitive level of accuracy for a large-scale and complex dataset, i.e., the COCO dataset.

Concept drift is a significant challenge for malware detection, as the performance of trained machine learning models degrades over time, rendering them impractical. While prior research in malware concept drift adaptation has primarily focused on active learning, which involves selecting representative samples to update the model, self-training has emerged as a promising approach to mitigate concept drift. Self-training involves retraining the model using pseudo labels to adapt to shifting data distributions. In this research, we propose MORPH -- an effective pseudo-label-based concept drift adaptation method specifically designed for neural networks. Through extensive experimental analysis of Android and Windows malware datasets, we demonstrate the efficacy of our approach in mitigating the impact of concept drift. Our method offers the advantage of reducing annotation efforts when combined with active learning. Furthermore, our method significantly improves over existing works in automated concept drift adaptation for malware detection.

Shapley values have been used extensively in machine learning, not only to explain black box machine learning models, but among other tasks, also to conduct model debugging, sensitivity and fairness analyses and to select important features for robust modelling and for further follow-up analyses. Shapley values satisfy certain axioms that promote fairness in distributing contributions of features toward prediction or reducing error, after accounting for non-linear relationships and interactions when complex machine learning models are employed. Recently, a number of feature selection methods utilising Shapley values have been introduced. Here, we present a novel feature selection method, LLpowershap, which makes use of loss-based Shapley values to identify informative features with minimal noise among the selected sets of features. Our simulation results show that LLpowershap not only identifies higher number of informative features but outputs fewer noise features compared to other state-of-the-art feature selection methods. Benchmarking results on four real-world datasets demonstrate higher or at par predictive performance of LLpowershap compared to other Shapley based wrapper methods, or filter methods.