Monitoring calf behaviour continuously would be beneficial to identify routine practices (e.g., weaning, dehorning, etc.) that impact calf welfare in dairy farms. In that regard, accelerometer data collected from neck collars can be used along with Machine Learning models to classify calf behaviour automatically. Hand-crafted features are commonly used in Machine Learning models, while ROCKET and Catch22 features are specifically designed for time-series classification problems in related fields. This study aims to compare the performance of ROCKET and Catch22 features to Hand-Crafted features. 30 Irish Holstein Friesian and Jersey pre-weaned calves were monitored using accelerometer sensors allowing for 27.4 hours of annotated behaviors. Additional time-series were computed from the raw X, Y and Z-axis and split into 3-second time windows. ROCKET, Catch22 and Hand-Crafted features were calculated for each time window, and the dataset was then split into the train, validation and test sets. Each set of features was used to train three Machine Learning models (Random Forest, eXtreme Gradient Boosting, and RidgeClassifierCV) to classify six behaviours indicative of pre-weaned calf welfare (drinking milk, grooming, lying, running, walking and other). Models were tuned with the validation set, and the performance of each feature-model combination was evaluated with the test set. The best performance across the three models was obtained with ROCKET [average balanced accuracy +/- standard deviation] (0.70 +/- 0.07), followed by Catch22 (0.69 +/- 0.05), surpassing Hand-Crafted (0.65 +/- 0.034). The best balanced accuracy (0.77) was obtained with ROCKET and RidgeClassifierCV, followed by Catch22 and Random Forest (0.73). Thus, tailoring these approaches for specific behaviours and contexts will be crucial in advancing precision livestock farming and enhancing animal welfare on a larger scale.

Missing value imputation is a crucial preprocessing step for many machine learning problems. However, it is often considered as a separate subtask from downstream applications such as classification, regression, or clustering, and thus is not optimized together with them. We hypothesize that treating the imputation model and downstream task model together and optimizing over full pipelines will yield better results than treating them separately. Our work describes a novel AutoML technique for making downstream predictions with missing data that automatically handles preprocessing, model weighting, and selection during inference time, with minimal compute overhead. Specifically we develop M-DEW, a Dynamic missingness-aware Ensemble Weighting (DEW) approach, that constructs a set of two-stage imputation-prediction pipelines, trains each component separately, and dynamically calculates a set of pipeline weights for each sample during inference time. We thus extend previous work on dynamic ensemble weighting to handle missing data at the level of full imputation-prediction pipelines, improving performance and calibration on downstream machine learning tasks over standard model averaging techniques. M-DEW is shown to outperform the state-of-the-art in that it produces statistically significant reductions in model perplexity in 17 out of 18 experiments, while improving average precision in 13 out of 18 experiments.

We present a multimodal traffic light state detection using vision and sound, from the viewpoint of a quadruped robot navigating in urban settings. This is a challenging problem because of the visual occlusions and noise from robot locomotion. Our method combines features from raw audio with the ratios of red and green pixels within bounding boxes, identified by established vision-based detectors. The fusion method aggregates features across multiple frames in a given timeframe, increasing robustness and adaptability. Results show that our approach effectively addresses the challenge of visual occlusion and surpasses the performance of single-modality solutions when the robot is in motion. This study serves as a proof of concept, highlighting the significant, yet often overlooked, potential of multi-modal perception in robotics.

Optimization techniques in deep learning are predominantly led by first-order gradient methodologies, such as SGD. However, neural network training can greatly benefit from the rapid convergence characteristics of second-order optimization. Newton's GD stands out in this category, by rescaling the gradient using the inverse Hessian. Nevertheless, one of its major bottlenecks is matrix inversion, which is notably time-consuming in $O(N^3)$ time with weak scalability. Matrix inversion can be translated into solving a series of linear equations. Given that quantum linear solver algorithms (QLSAs), leveraging the principles of quantum superposition and entanglement, can operate within a $\text{polylog}(N)$ time frame, they present a promising approach with exponential acceleration. Specifically, one of the most recent QLSAs demonstrates a complexity scaling of $O(d\cdot\kappa \log(N\cdot\kappa/\epsilon))$, depending on: {size~$N$, condition number~$\kappa$, error tolerance~$\epsilon$, quantum oracle sparsity~$d$} of the matrix. However, this also implies that their potential exponential advantage may be hindered by certain properties (i.e. $\kappa$ and $d$). We propose Q-Newton, a hybrid quantum-classical scheduler for accelerating neural network training with Newton's GD. Q-Newton utilizes a streamlined scheduling module that coordinates between quantum and classical linear solvers, by estimating & reducing $\kappa$ and constructing $d$ for the quantum solver. Our evaluation showcases the potential for Q-Newton to significantly reduce the total training time compared to commonly used optimizers like SGD. We hypothesize a future scenario where the gate time of quantum machines is reduced, possibly realized by attoseconds physics. Our evaluation establishes an ambitious and promising target for the evolution of quantum computing.

Hierarchy Of Time-Surfaces is a neuromorphic algorithm used to extract features from patterns of events [1]. This is possible thanks to a type of representation called time-surface or time vector, where events are interpolated by exponential decay kernels and collected to represent relative time differences between the activation of units in the network. Time surfaces are one of the most common representations in the neuromorphic field since they allow to interface event data with traditional machine learning and computer vision algorithms [2, 3]. In HOTS, time surfaces are clustered together using algorithms like k-means to extract common activity patterns, and layers of units are built by considering each centroid as a neuron that can emit a new event when an input time surface is assigned to it. For this reason, HOTS shares many points in common with bag-of-words or bag-of-features algorithms[4]. For instance, HOTS requires an external classifier on histograms of features to classify information. Similarly to bag-of-words algorithms, HOTS classifiers are histograms that accumulate features over a given temporal window to produce an input vector to traditional machine learning algorithms like Support Vector Machines and Multi-Layer Perceptrons[5, 6, 1, 7]. This approach limits compatibility with neuromorphic hardware and can nullify latency and energy efficiency advantages that are found in neuromorphic systems. Compared to Spiking Neural Networks (SNNs) trained with backpropagation through time, HOTS lags in accuracy [8, 9, 10, 1].

Parkinson's disease is a widespread neurodegenerative condition necessitating early diagnosis for effective intervention. This paper introduces an innovative method for diagnosing Parkinson's disease through the analysis of human EEG signals, employing a Support Vector Machine (SVM) classification model. this research presents novel contributions to enhance diagnostic accuracy and reliability. Our approach incorporates a comprehensive review of EEG signal analysis techniques and machine learning methods. Drawing from recent studies, we have engineered an advanced SVM-based model optimized for Parkinson's disease diagnosis. Utilizing cutting-edge feature engineering, extensive hyperparameter tuning, and kernel selection, our method achieves not only heightened diagnostic accuracy but also emphasizes model interpretability, catering to both clinicians and researchers. Moreover, ethical concerns in healthcare machine learning, such as data privacy and biases, are conscientiously addressed. We assess our method's performance through experiments on a diverse dataset comprising EEG recordings from Parkinson's disease patients and healthy controls, demonstrating significantly improved diagnostic accuracy compared to conventional techniques. In conclusion, this paper introduces an innovative SVM-based approach for diagnosing Parkinson's disease from human EEG signals. Building upon the IEEE framework and previous research, its novelty lies in the capacity to enhance diagnostic accuracy while upholding interpretability and ethical considerations for practical healthcare applications. These advances promise to revolutionize early Parkinson's disease detection and management, ultimately contributing to enhanced patient outcomes and quality of life.

In the ever-evolving landscape of machine learning, seamless translation of natural language descriptions into executable code remains a formidable challenge. This paper introduces Linguacodus, an innovative framework designed to tackle this challenge by deploying a dynamic pipeline that iteratively transforms natural language task descriptions into code through high-level data-shaping instructions. The core of Linguacodus is a fine-tuned large language model (LLM), empowered to evaluate diverse solutions for various problems and select the most fitting one for a given task. This paper details the fine-tuning process, and sheds light on how natural language descriptions can be translated into functional code. Linguacodus represents a substantial leap towards automated code generation, effectively bridging the gap between task descriptions and executable code. It holds great promise for advancing machine learning applications across diverse domains. Additionally, we propose an algorithm capable of transforming a natural description of an ML task into code with minimal human interaction. In extensive experiments on a vast machine learning code dataset originating from Kaggle, we showcase the effectiveness of Linguacodus. The investigations highlight its potential applications across diverse domains, emphasizing its impact on applied machine learning in various scientific fields.

Object detectors are typically trained once and for all on a fixed set of classes. However, this closed-world assumption is unrealistic in practice, as new classes will inevitably emerge after the detector is deployed in the wild. In this work, we look at ways to extend a detector trained for a set of base classes so it can i) spot the presence of novel classes, and ii) automatically enrich its repertoire to be able to detect those newly discovered classes together with the base ones. We propose PANDAS, a method for novel class discovery and detection. It discovers clusters representing novel classes from unlabeled data, and represents old and new classes with prototypes. During inference, a distance-based classifier uses these prototypes to assign a label to each detected object instance. The simplicity of our method makes it widely applicable. We experimentally demonstrate the effectiveness of PANDAS on the VOC 2012 and COCO-to-LVIS benchmarks. It performs favorably against the state of the art for this task while being computationally more affordable.

We study overparameterization in generative adversarial networks (GANs) that can interpolate the training data. We show that overparameterization can improve generalization performance and accelerate the training process. We study the generalization error as a function of latent space dimension and identify two main behaviors, depending on the learning setting. First, we show that overparameterized generative models that learn distributions by minimizing a metric or $f$-divergence do not exhibit double descent in generalization errors; specifically, all the interpolating solutions achieve the same generalization error. Second, we develop a novel pseudo-supervised learning approach for GANs where the training utilizes pairs of fabricated (noise) inputs in conjunction with real output samples. Our pseudo-supervised setting exhibits double descent (and in some cases, triple descent) of generalization errors. We combine pseudo-supervision with overparameterization (i.e., overly large latent space dimension) to accelerate training while matching or even surpassing generalization performance without pseudo-supervision. While our analysis focuses mostly on linear models, we also apply important insights for improving generalization of nonlinear, multilayer GANs.

Brain responses related to working memory originate from distinct brain areas and oscillate at different frequencies. EEG signals with high temporal correlation can effectively capture these responses. Therefore, estimating the functional connectivity of EEG for working memory protocols in different frequency bands plays a significant role in analyzing the brain dynamics with increasing memory and cognitive loads, which remains largely unexplored. The present study introduces a Bayesian structure learning algorithm to learn the functional connectivity of EEG in sensor space. Next, the functional connectivity graphs are taken as input to the graph convolutional network to classify the working memory loads. The intrasubject (subject-specific) classification performed on 154 subjects for six different verbal working memory loads produced the highest classification accuracy of 96% and average classification accuracy of 89%, outperforming state-of-the-art classification models proposed in the literature. Furthermore, the proposed Bayesian structure learning algorithm is compared with state-of-the-art functional connectivity estimation methods through intersubject and intrasubject statistical analysis of variance. The results also show that the alpha and theta bands have better classification accuracy than the beta band.