Low-Earth orbit (LEO) satellites have been prosperously deployed for various Earth observation missions due to its capability of collecting a large amount of image or sensor data. However, traditionally, the data training process is performed in the terrestrial cloud server, which leads to a high transmission overhead. With the recent development of LEO, it is more imperative to provide ultra-dense LEO constellation with enhanced on-board computation capability. Benefited from it, we have proposed a collaborative federated learning for low Earth orbit (FELLO). We allocate the entire process on LEOs with low payload inter-satellite transmissions, whilst the low-delay terrestrial gateway server (GS) only takes care for initial signal controlling. The GS initially selects an LEO server, whereas its LEO clients are all determined by clustering mechanism and communication capability through the optical inter-satellite links (ISLs). The re-clustering of changing LEO server will be executed once with low communication quality of FELLO. In the simulations, we have numerically analyzed the proposed FELLO under practical Walker-based LEO constellation configurations along with MNIST training dataset for classification mission. The proposed FELLO outperforms the conventional centralized and distributed architectures with higher classification accuracy as well as comparably lower latency of joint communication and computing.

Sentiment analysis is the process of identifying and extracting subjective information from text. Despite the advances to employ cross-lingual approaches in an automatic way, the implementation and evaluation of sentiment analysis systems require language-specific data to consider various sociocultural and linguistic peculiarities. In this paper, the collection and annotation of a dataset are described for sentiment analysis of Central Kurdish. We explore a few classical machine learning and neural network-based techniques for this task. Additionally, we employ an approach in transfer learning to leverage pretrained models for data augmentation. We demonstrate that data augmentation achieves a high F$_1$ score and accuracy despite the difficulty of the task.

Are we alone in the universe? Scientists may have just moved us closer to answering this question. The team – led by researchers from the University of Toronto – has streamlined the search for extraterrestrial life by using a new algorithm to organize the data from their telescopes into categories, in order to distinguish between real signals and interference. This has allowed them to quickly sort through the information and find patterns, through an artificial intelligence process known as machine learning. They discovered eight extraterrestrial signals that seem to have the hallmarks of technology.

Black-box adversarial attacks have shown strong potential to subvert machine learning models. Existing black-box adversarial attacks craft the adversarial examples by iteratively querying the target model and/or leveraging the transferability of a local surrogate model. Whether such attack can succeed remains unknown to the adversary when empirically designing the attack. In this paper, to our best knowledge, we take the first step to study a new paradigm of adversarial attacks -- certifiable black-box attack that can guarantee the attack success rate of the crafted adversarial examples. Specifically, we revise the randomized smoothing to establish novel theories for ensuring the attack success rate of the adversarial examples. To craft the adversarial examples with the certifiable attack success rate (CASR) guarantee, we design several novel techniques, including a randomized query method to query the target model, an initialization method with smoothed self-supervised perturbation to derive certifiable adversarial examples, and a geometric shifting method to reduce the perturbation size of the certifiable adversarial examples for better imperceptibility. We have comprehensively evaluated the performance of the certifiable black-box attack on CIFAR10 and ImageNet datasets against different levels of defenses. Both theoretical and experimental results have validated the effectiveness of the proposed certifiable attack.

In recent years, graph representation learning techniques have proven effective in discovering meaningful vector representations of nodes, edges, or entire graphs, resulting in successful applications across a wide range of downstream tasks [29, 52, 68]. However, graph data often presents a significant challenge in the form of class imbalance, where one class's instances significantly outnumber those of other classes. This imbalance can lead to suboptimal performance when applying machine learning techniques to graph data. Class-imbalanced learning on graphs (CILG) is an emerging research area addressing class imbalance in graph data, where traditional methods for non-graph data might be unsuitable or ineffective for several reasons. Firstly, graph data's unique, irregular, non-Euclidean structure complicates traditional class-imbalance techniques designed for Euclidean data [78]. Secondly, graph data often holds rich relational information, necessitating specialized techniques for preservation and leverage during the learning process [51]. Lastly, node dependencies and interactions in a graph make class re-balancing complex, as naïve oversampling or undersampling may disrupt the graph's structure and thus lead to poor performance [35].

Decentralized learning enables serverless training of deep neural networks (DNNs) in a distributed manner on multiple nodes. This allows for the use of large datasets, as well as the ability to train with a wide variety of data sources. However, one of the key challenges with decentralized learning is heterogeneity in the data distribution across the nodes. In this paper, we propose In-Distribution Knowledge Distillation (IDKD) to address the challenge of heterogeneous data distribution. The goal of IDKD is to homogenize the data distribution across the nodes. While such data homogenization can be achieved by exchanging data among the nodes sacrificing privacy, IDKD achieves the same objective using a common public dataset across nodes without breaking the privacy constraint. This public dataset is different from the training dataset and is used to distill the knowledge from each node and communicate it to its neighbors through the generated labels. With traditional knowledge distillation, the generalization of the distilled model is reduced because all the public dataset samples are used irrespective of their similarity to the local dataset. Thus, we introduce an Out-of-Distribution (OoD) detector at each node to label a subset of the public dataset that maps close to the local training data distribution. Finally, only labels corresponding to these subsets are exchanged among the nodes and with appropriate label averaging each node is finetuned on these data subsets along with its local data. Our experiments on multiple image classification datasets and graph topologies show that the proposed IDKD scheme is more effective than traditional knowledge distillation and achieves state-of-the-art generalization performance on heterogeneously distributed data with minimal communication overhead.

Updated Turnitin, which touts itself as a maker of plagiarism-detecting software for academia, demoed on Tuesday what we're told is a tool capable of detecting AI writing. Crucially, it boasted its machine learning software will flag up computer-generated cheating when it has at least "98 percent confidence" it is right. Be as that may, that's not the same as accuracy. A model can be supremely confident and still completely wrong most of the time. Indeed, tests carried out by the Washington Post found the detection tool's performance lacking.

As per estimation, 8 million people could not get Consider a post in a social media platform - Reddit specialist help as they were not considered sick enough to posted in subreddit r/depression. The post is personally qualify. This situation underscores the need for automation written by a user which exhibit higher levels of emotions of mental health detection from social media data where and stances associated with mental and social well-being, people express themselves and their thoughts, beliefs/ emotions respectively. The post written by user is given as: with ease. This self-reported social media data is valuable but laborious for manual interpretations; thus, although = "I do not want to read literature but my complex, an automated system would significantly enhance parents forced me to do so. Not happy with my the ability to understand a social media user's state of mental grades" health. Amid COVID-19, the social NLP research community A major concern of user is about education stating the witness increase in the use of social media to express issue of forced subjects by parents that clearly indicate thoughts/ feelings and share life experiences Gianfredi, child's lack of interest affecting state of mind.

In the task of comparing two classification algorithms, the widely-used McNemar's test aims to infer the presence of a significant difference between the error rates of the two classification algorithms. However, the power of the conventional McNemar's test is usually unpromising because the hold-out (HO) method in the test merely uses a single train-validation split that usually produces a highly varied estimation of the error rates. In contrast, a cross-validation (CV) method repeats the HO method in multiple times and produces a stable estimation. Therefore, a CV method has a great advantage to improve the power of McNemar's test. Among all types of CV methods, a block-regularized 5$\times$2 CV (BCV) has been shown in many previous studies to be superior to the other CV methods in the comparison task of algorithms because the 5$\times$2 BCV can produce a high-quality estimator of the error rate by regularizing the numbers of overlapping records between all training sets. In this study, we compress the 10 correlated contingency tables in the 5$\times$2 BCV to form an effective contingency table. Then, we define a 5$\times$2 BCV McNemar's test on the basis of the effective contingency table. We demonstrate the reasonable type I error and the promising power of the proposed 5$\times$2 BCV McNemar's test on multiple simulated and real-world data sets.

Out-of-distribution (OOD) detection has received much attention lately due to its practical importance in enhancing the safe deployment of neural networks. One of the primary challenges is that models often produce highly confident predictions on OOD data, which undermines the driving principle in OOD detection that the model should only be confident about in-distribution samples. In this work, we propose ReAct--a simple and effective technique for reducing model overconfidence on OOD data. Our method is motivated by novel analysis on internal activations of neural networks, which displays highly distinctive signature patterns for OOD distributions. Our method can generalize effectively to different network architectures and different OOD detection scores. We empirically demonstrate that ReAct achieves competitive detection performance on a comprehensive suite of benchmark datasets, and give theoretical explication for our method's efficacy. On the ImageNet benchmark, ReAct reduces the false positive rate (FPR95) by 25.05% compared to the previous best method