Search

In recent years, financial analysts have been trying to develop models to predict the movement of a stock price index. The task becomes challenging in vague economic, social, and political situations like in Pakistan. In this study, we employed efficient models of machine learning such as long short-term memory (LSTM) and quantum long short-term memory (QLSTM) to predict the Karachi Stock Exchange (KSE) 100 index by taking monthly data of twenty-six economic, social, political, and administrative indicators from February 2004 to December 2020. The comparative results of LSTM and QLSTM predicted values of the KSE 100 index with the actual values suggested QLSTM a potential technique to predict stock market trends.

Environmental Sound Classification is an important problem of sound recognition and is more complicated than speech recognition problems as environmental sounds are not well structured with respect to time and frequency. Researchers have used various CNN models to learn audio features from different audio features like log mel spectrograms, gammatone spectral coefficients, mel-frequency spectral coefficients, generated from the audio files, over the past years. In this paper, we propose a new methodology : Two-Level Classification; the Level 1 Classifier will be responsible to classify the audio signal into a broader class and the Level 2 Classifiers will be responsible to find the actual class to which the audio belongs, based on the output of the Level 1 Classifier. We have also shown the effects of different audio filters, among which a new method of Audio Crop is introduced in this paper, which gave the highest accuracies in most of the cases. We have used the ESC-50 dataset for our experiment and obtained a maximum accuracy of 78.75% in case of Level 1 Classification and 98.04% in case of Level 2 Classifications., Comment: 19 pages, 16 figures

A language is made up of an infinite/finite number of sentences, which in turn is composed of a number of words. The Electrocardiogram (ECG) is the most popular noninvasive medical tool for studying heart function and diagnosing various irregular cardiac rhythms. Intuitive inspection of the ECG reveals a marked similarity between ECG signals and the spoken language. As a result, the ECG signal may be thought of as a series of heartbeats (similar to sentences in a spoken language), with each heartbeat consisting of a collection of waves (similar to words in a sentence) with varying morphologies. Just as natural language processing (NLP) is used to help computers comprehend and interpret human natural language, it is conceivable to create NLP-inspired algorithms to help computers comprehend the electrocardiogram data more efficiently. In this study, we propose a novel ECG analysis technique, based on embedding and self attention, to capture the spatial as well as the temporal dependencies of the ECG data. To generate the embedding, an encoder-decoder network was proposed to capture the temporal dependencies of the ECG signal and perform data compression. The compressed and encoded data was fed to the embedding layer as its weights. Finally, the proposed CNN-LSTM-Self Attention classifier works on the embedding layer and classifies the signal as normal or anomalous. The approach was tested using the PTB-xl dataset, which is severely imbalanced. Our emphasis was to appropriately recognise the disease classes present in minority numbers, in order to limit the detection of False Negative cases. An accuracy of 91% was achieved with a good F1-score for all the disease classes. Additionally, the the size of the model was reduced by 34% due to compression, making it suitable for deployment in real time applications

There have been significant issues given the IoT, with heterogeneity of billions of devices and with a large amount of data. This paper proposed an innovative design of the Internet of Things (IoT) Environment Intrusion Detection System (or IDS) using Deep Learning-integrated Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. Our model, based on the CICIDS2017 dataset, achieved an accuracy of 99.52% in classifying network traffic as either benign or malicious. The real-time processing capability, scalability, and low false alarm rate in our model surpass some traditional IDS approaches and, therefore, prove successful for application in today's IoT networks. The development and the performance of the model, with possible applications that may extend to other related fields of adaptive learning techniques and cross-domain applicability, are discussed. The research involving deep learning for IoT cybersecurity offers a potent solution for significantly improving network security.

One of the principal objectives of Natural Language Processing (NLP) is to generate meaningful representations from text. Improving the informativeness of the representations has led to a tremendous rise in the dimensionality and the memory footprint. It leads to a cascading effect amplifying the complexity of the downstream model by increasing its parameters. The available techniques cannot be applied to cross-modal applications such as text-to-image. To ameliorate these issues, a novel Text-to-Image methodology for generating fixed-length representations through a self-supervised Variational Auto-Encoder (VAE) for semantic evaluation applying transformers (TexIm FAST) has been proposed in this paper. The pictorial representations allow oblivious inference while retaining the linguistic intricacies, and are potent in cross-modal applications. TexIm FAST deals with variable-length sequences and generates fixed-length representations with over 75% reduced memory footprint. It enhances the efficiency of the models for downstream tasks by reducing its parameters. The efficacy of TexIm FAST has been extensively analyzed for the task of Semantic Textual Similarity (STS) upon the MSRPC, CNN/ Daily Mail, and XSum data-sets. The results demonstrate 6% improvement in accuracy compared to the baseline and showcase its exceptional ability to compare disparate length sequences such as a text with its summary., Comment: 19 pages, 33 figures

The advent of transformers with attention mechanisms and associated pre-trained models have revolutionized the field of Natural Language Processing (NLP). However, such models are resource-intensive due to highly complex architecture. This limits their application to resource-constrained environments. While choosing an appropriate NLP model, a major trade-off exists over choosing accuracy over efficiency and vice versa. This paper presents a commentary on the evolution of NLP and its applications with emphasis on their accuracy as-well-as efficiency. Following this, a survey of research contributions towards enhancing the efficiency of transformer-based models at various stages of model development along with hardware considerations has been conducted. The goal of this survey is to determine how current NLP techniques contribute towards a sustainable society and to establish a foundation for future research.

Available training data for named entity recognition (NER) often contains a significant percentage of incorrect labels for entity types and entity boundaries. Such label noise poses challenges for supervised learning and may significantly deteriorate model quality. To address this, prior work proposed various noise-robust learning approaches capable of learning from data with partially incorrect labels. These approaches are typically evaluated using simulated noise where the labels in a clean dataset are automatically corrupted. However, as we show in this paper, this leads to unrealistic noise that is far easier to handle than real noise caused by human error or semi-automatic annotation. To enable the study of the impact of various types of real noise, we introduce NoiseBench, an NER benchmark consisting of clean training data corrupted with 6 types of real noise, including expert errors, crowdsourcing errors, automatic annotation errors and LLM errors. We present an analysis that shows that real noise is significantly more challenging than simulated noise, and show that current state-of-the-art models for noise-robust learning fall far short of their theoretically achievable upper bound. We release NoiseBench to the research community., Comment: data available at https://github.com/elenamer/NoiseBench; to appear at EMNLP2024 main conference

The safe transition from conditional automation to manual driving control is significantly intertwined with the vehicle's lateral and longitudinal dynamics. The transition may occur as a result of a system-initiated mandatory takeover (MTOR) or as a driver-initiated discretionary takeover (DTOR). In either condition, the takeover process entails differing cognitive demands and may affect the driving behaviour differently. This study analyzes driving stability and perceived mental workload in 304 takeover attempts recorded from 104 participants within virtual and immersive reality environments. Adopting an exploratory approach, this dynamic simulator study employs a mixed factorial design. Utilizing a deep neural network-based survival analysis with SHAP interpretability, the study investigated the influence of covariates on perception-reaction time (PRT), distinguishing between safe and unsafe control transition and offering insights into the temporal dynamics of these shifts. The distributions of key parameters in experimental groups were analyzed and factors influencing the perceived mental workload were estimated using multivariate linear regression. The findings indicate a notable decrease in the risk of unsafe takeovers (described by a longer PRT) when drivers have prior control-transition experience and familiarity with Automated Vehicles (AVs). However, driver's prior familiarity and experience with AVs only decreased the perceived mental workload associated with DTOR, with an insignificant impact on the cognitive demand of MTOR. Furthermore, multitasking during automated driving significantly elevated the cognitive demand linked to DTOR and led to longer PRT in MTOR situations.

Instruction-tuned Large Language Models (LLMs) have recently showcased remarkable advancements in their ability to generate fitting responses to natural language instructions. However, many current works rely on manual evaluation to judge the quality of generated responses. Since such manual evaluation is time-consuming, it does not easily scale to the evaluation of multiple models and model variants. In this short paper, we propose a straightforward but remarkably effective evaluation metric called SemScore, in which we directly compare model outputs to gold target responses using semantic textual similarity (STS). We conduct a comparative evaluation of the model outputs of 12 prominent instruction-tuned LLMs using 8 widely-used evaluation metrics for text generation. We find that our proposed SemScore metric outperforms all other, in many cases more complex, evaluation metrics in terms of correlation to human evaluation. These findings indicate the utility of our proposed metric for the evaluation of instruction-tuned LLMs.

Sampling autonomy for icy moon lander missions requires understanding of topographic and photometric properties of the sampling terrain. Unavailability of high resolution visual datasets (either bird-eye view or point-of-view from a lander) is an obstacle for selection, verification or development of perception systems. We attempt to alleviate this problem by: 1) proposing Graphical Utility for Icy moon Surface Simulations (GUISS) framework, for versatile stereo dataset generation that spans the spectrum of bulk photometric properties, and 2) focusing on a stereo-based visual perception system and evaluating both traditional and deep learning-based algorithms for depth estimation from stereo matching. The surface reflectance properties of icy moon terrains (Enceladus and Europa) are inferred from multispectral datasets of previous missions. With procedural terrain generation and physically valid illumination sources, our framework can fit a wide range of hypotheses with respect to visual representations of icy moon terrains. This is followed by a study over the performance of stereo matching algorithms under different visual hypotheses. Finally, we emphasize the standing challenges to be addressed for simulating perception data assets for icy moons such as Enceladus and Europa. Our code can be found here: https://github.com/nasa-jpl/guiss., Comment: Software: https://github.com/nasa-jpl/guiss. IEEE Aerospace Conference 2024

This paper proposes the incorporation of techniques from stereophotoclinometry (SPC) into a keypoint-based structure-from-motion (SfM) system to estimate the surface normal and albedo at detected landmarks to improve autonomous surface and shape characterization of small celestial bodies from in-situ imagery. In contrast to the current state-of-the-practice method for small body shape reconstruction, i.e., SPC, which relies on human-in-the-loop verification and high-fidelity a priori information to achieve accurate results, we forego the expensive maplet estimation step and instead leverage dense keypoint measurements and correspondences from an autonomous keypoint detection and matching method based on deep learning to provide the necessary photogrammetric constraints. Moreover, we develop a factor graph-based approach allowing for simultaneous optimization of the spacecraft's pose, landmark positions, Sun-relative direction, and surface normals and albedos via fusion of Sun sensor measurements and image keypoint measurements. The proposed framework is validated on real imagery of the Cornelia crater on Asteroid 4 Vesta, along with pose estimation and mapping comparison against an SPC reconstruction, where we demonstrate precise alignment to the SPC solution without relying on any a priori camera pose and topography information or humans-in-the-loop

Stellar bars in disk galaxies grow as stars in near circular orbits lose angular momentum to their environments, including their Dark Matter (DM) halo, and transform into elongated bar orbits. This angular momentum exchange during galaxy evolution hints at a connection between bar properties and the DM halo spin $\lambda$, the dimensionless form of DM angular momentum. We investigate the connection between halo spin $\lambda$ and galaxy properties in the presence/absence of stellar bars, using the cosmological magneto-hydrodynamic TNG50 simulations at multiple redshifts ($00.4$) reside in DM halos with low spin and low specific angular momentum at their centers. In contrast, unbarred/weakly barred galaxies ($A_2/A_0<0.2$) exist in halos with higher central spin and higher specific angular momentum. The anti-correlation is due to the barred galaxies' higher DM mass and lower angular momentum than the unbarred galaxies at $z_r=0$, as a result of galaxy evolution. At high redshifts ($z_r=1$), all galaxies have higher halo spin compared to those at lower redshifts ($z_r=0$), with a weak anti-correlation for galaxies having $A_2/A_0> 0.2$. The formation of a DM bar in strongly barred systems highlights how angular momentum transfer to the halo can influence its central spin., Comment: 24 pages (including Appendix), 15 figures; accepted in ApJ; comments are welcome!

Recently, canted antiferromagnets offer great potential for fundamental research and applications due to their unique properties. The presence of the Dzyaloshinskii-Moriya interaction leads to the existence of a weak ferromagnetic moment at room temperature. We study both theoretically and experimentally microwave spin pumping by the quasi-ferromagnetic mode from a canted easy plane antiferromagnet with weak ferromagnetism FeBO3. The conversion of a microwave signal into the constant voltage is realized using the inverse spin Hall effect in an iron borate/heavy metal heterostructure. We use an additional bias magnetic field to selectively tune the resonance frequency of such a microwave detector over a wide range up to 43.5 GHz with potential sensitivity near 2.5 microV/W. We confirm the pure spin current nature by changing polarity of the detected via inverse spin Hall effect voltage by switching the direction of the bias magnetic field. We believe that our results will be useful for the development of highly tunable, portable and sensitive microwave antiferromagnet-based functional devices., Comment: 7 pages, 3 figures, 3 pages supplementary information

The physical mechanisms responsible for bar formation and destruction in galaxies remain a subject of debate. While we have gained valuable insight into how bars form and evolve from isolated idealized simulations, in the cosmological domain, galactic bars evolve in complex environments with mergers, gas accretion events, in presence of turbulent Inter Stellar Medium (ISM) with multiple star formation episodes, in addition to coupling to their host galaxies' dark matter halos. We investigate bar formation in 13 Milky Way-mass galaxies from the FIRE-2 (Feedback in Realistic Environments) cosmological zoom-in simulations. 8 of the 13 simulated galaxies form bars at some point during their history: three from tidal interactions and five from internal evolution of the disk. The bars in FIRE-2 are generally shorter than the corotation radius (mean bar radius $\sim 1.53$ kpc), have a wide range of pattern speeds (36--97 km s$^{-1}$kpc$^{-1}$), and live for a wide range of dynamical times (2--160 bar rotations). We find that bar formation in FIRE-2 galaxies is influenced by satellite interactions and the stellar-to-dark matter mass ratio in the inner galaxy, but neither is a sufficient condition for bar formation. Bar formation is more likely to occur, and the bars formed are stronger and longer-lived, if the disks are kinematically cold; galaxies with high central gas fractions and/or vigorous star formation, on the other hand, tend to form weaker bars. In the case of the FIRE-2 galaxies these properties combine to produce ellipsoidal bars with strengths $A_2/A_0 \sim$ 0.1--0.2., Comment: 39 pages (including Appendix), 25 figures, accepted in The Astrophysical Journal

Instruction-tuned Large Language Models (LLMs) have recently showcased remarkable ability to generate fitting responses to natural language instructions. However, an open research question concerns the inherent biases of trained models and their responses. For instance, if the data used to tune an LLM is dominantly written by persons with a specific political bias, we might expect generated answers to share this bias. Current research work seeks to de-bias such models, or suppress potentially biased answers. With this demonstration, we take a different view on biases in instruction-tuning: Rather than aiming to suppress them, we aim to make them explicit and transparent. To this end, we present OpinionGPT, a web demo in which users can ask questions and select all biases they wish to investigate. The demo will answer this question using a model fine-tuned on text representing each of the selected biases, allowing side-by-side comparison. To train the underlying model, we identified 11 different biases (political, geographic, gender, age) and derived an instruction-tuning corpus in which each answer was written by members of one of these demographics. This paper presents OpinionGPT, illustrates how we trained the bias-aware model and showcases the web application (available at https://opiniongpt.informatik.hu-berlin.de)., Comment: 6 pages, 1 figure, 3 tables

The driver's willingness to give (WTG) control in conditionally automated driving is assessed in a virtual reality based driving-rig, through their choice to give away driving control and through the extent to which automated driving is adopted in a mixed-traffic environment. Within- and across-class unobserved heterogeneity and locus of control variations are taken into account. The choice of giving away control is modelled using the mixed logit (MIXL) and mixed latent class (LCML) model. The significant latent segments of the locus of control are developed into internalizers and externalizers by the latent class model (LCM) based on the taste heterogeneity identified from the MIXL model. Results suggest that drivers choose to "giveAway" control of the vehicle when greater concentration/attentiveness is required (e.g., in the nighttime) or when they are interested in performing a non-driving-related task (NDRT). In addition, it is observed that internalizers demonstrate more heterogeneity compared to externalizers in terms of WTG.

Dark matter (DM) halo angular momentum is very challenging to determine from observations of galaxies. In this study, we present a new hybrid method of estimating the dimensionless halo angular momentum, halo spin of a gas-rich dwarf barred galaxy UGC5288 using N-Body/SPH simulations. We forward model the galaxy disk properties: stellar and gas mass, surface densities, disk scalelengths, bar length and bar ellipticity from observations. We use the HI rotation curve to constrain the DM halo density profile and further use the bar properties to determine the models that best represent the observed baryonic disk. We compare the halo spin profile from our models to the halo spin profiles of similar mass dwarf galaxy analogues of UGC5288 in the TNG50 simulations. The halo spin profile from our simulated models matches within ballpark values of the median spin profile of UGC5288 analogues in the TNG50 simulations, although there are some uncertainties due to the DM halo evolutionary history., Comment: 4 pages, 3 figures, part of Dynamical Masses of Local Group Galaxies: IAU Symposium 379

Dark matter (DM) halo properties are extensively studied in cosmological simulations but are very challenging to estimate from observations. The DM halo density profile of observed galaxies is modelled using multiple probes that trace the dark matter potential. However, the angular momentum distribution of DM halos is still a subject of debate. In this study we investigate a method for estimating the halo spin and halo concentration of low surface brightness (LSB), gas-rich dwarf barred galaxy UGC 5288, by forward modelling disk properties derived from observations - stellar and gas surface densities, disk scale length, HI rotation curve, bar length and bar ellipticity. We combine semi-analytical techniques, N-body/SPH and cosmological simulations to model the DM halo of UGC 5288 with both a cuspy Hernquist profile and a flat-core pseudo-isothermal profile. We find that the best match with observations is a pseudo-isothermal halo model with a core radius of $r_{c} = 0.23$ kpc, and halo spin of $\lambda$= 0.08 at the virial radius. Although our findings are consistent with previous core radius estimates of the halo density profile of UGC 5288, as well as with the halo spin profiles of similar mass analogues of UGC5288 in the high-resolution cosmological-magneto-hydrodynamical simulation TNG50, there still remain some uncertainties as we are limited in our knowledge of the formation history of the galaxy. Additionally, we find that the inner halo spin ($ r< 10$ kpc) in barred galaxies is different from the unbarred ones, and the halo spin shows weak correlations with bar properties., Comment: 29 pages including Appendix, 28 figures, accepted in MNRAS, comments are welcome from the community