Optimizing neural network techniques in classifying Fermi-LAT gamma-ray sources

The Fermi Large Area Telescope (LAT) has detected 6659 γ-ray sources in the incremental version (4FGL-DR3, for Data Release 3) of the fourth Fermi-LAT catalog of γ-ray sources and 3743 of them are blazars, including 1517 blazar candidates of uncertain type (BCUs). Blazars are generally classified by properties of emission lines into BL Lac objects and flat spectrum radio quasars (FSRQs). However, BCUs are difficult to classify because of the lack of spectrum. In this work we apply five different machine-learning algorithms (K-nearest neighbors, logistic regression, support vector machine, random forest, CatBoost) to evaluate the classification of 1517 BCUs based on the observational data of 4FGL-DR3. The results indicate that the use of recursive feature elimination cross-validation can effectively improve the accuracy of models and reduce computation time. We use our models to predict the BCUs from 4FGL-DR3 and the results of the overlapping of the five models are as follows: 811 BL Lac objects, 397 FSRQs, and 309 BCUs.

While molecular dynamics (MD) is a very useful computational method for atomistic simulations, modeling the interatomic interactions for reliable MD simulations of real materials has been a long-standing challenge. In 2007, Behler and Parrinello first proposed and demonstrated an artificial neural network machine learning (ANN-ML) scheme, opening a new paradigm for developing accurate and efficient interatomic potentials for reliable MD simulation studies of the thermodynamics and kinetics of materials. In this paper, we show that an accurate and transferable ANN-ML interatomic potential can be developed for MD simulations of the La-Si-P system. The crucial role of training data in the ML potential development is discussed. The developed ANN-ML potential accurately describes not only the energy vs volume curves for all the known elemental, binary, and ternary crystalline structures in the La-Si-P system but also the structures of La-Si-P liquids with various compositions. Using the developed ANN-ML potential, the melting temperatures of several crystalline phases in the La-Si-P system are predicted by the coexistence of solid-liquid phases from MD simulations. While the ANN-ML model systematically underestimates the melting temperatures of these phases, the overall trend agrees with experiment. The developed ANN-ML potential is also applied to study the nucleation and growth of LaP as a function of different relative concentrations of Si and P in the La-Si-P liquid, and the obtained results are consistent with experimental observations.

Blazars, one of the most extreme classes of active galaxies, constitute so far the largest known population of γ-ray sources, and their number is continuously growing in the Fermi catalogs. However, in the latest release of the Fermi catalog there is still a large fraction of sources that are classified as blazar candidates of uncertain type (BCUs) for which optical spectroscopic observations are necessary to confirm their nature and their associations. In addition, about one-third of the γ-ray point sources listed in the Third Fermi-LAT Source Catalog (3FGL) are still unassociated and lacking an assigned lower-energy counterpart. Since 2012 we have been carrying out an optical spectroscopic campaign to observe blazar candidates to confirm their nature. In this paper, the sixth of the series, we present optical spectroscopic observations for 30 γ-ray blazar candidates from different observing programs we carried out with the Telescopio Nazionale Galileo, William Herschel Telescope, Observatorio Astronómico Nacional, Southern Astrophysical Research Telescope, and Magellan Telescopes. We found that 21 out of 30 sources investigated are BL Lac objects, while the remaining targets are classified as flat-spectrum radio quasars showing the typical broad emission lines of normal quasi-stellar objects. We conclude that our selection of γ-ray blazar candidates based on their multifrequency properties continues to be a successful way to discover potential low-energy counterparts of the Fermi unidentified gamma-ray sources and to confirm the nature of BCUs.

International audience

ABSTRACTThe deepest all-sky survey available in the γ-ray band – the last release of the Fermi-LAT catalogue (4FGL-DR3) based on the data accumulated in 12 years contains more than 6600 sources. The largest population among the sources is blazar subclass – 3743, 60.1 per cent of which are classified as BL Lacertae objects (BL Lacs) or Flat Spectrum Radio Quasars (FSRQs), while the rest are listed as blazar candidates of uncertain type (BCU) as their firm optical classification is lacking. The goal of this study is to classify BCUs using different machine learning algorithms, which are trained on the spectral and temporal properties of already classified BL Lacs and FSRQs. Artificial Neural Networks, XGBoost, and LightGBM algorithms are employed to construct predictive models for BCU classification. Using 18 input parameters of 2219 BL Lacs and FSRQs, we train (80 per cent of the sample) and test (20 per cent) these algorithms and find that LightGBM model, state-of-the-art classification algorithm based on gradient boosting decision trees, provides the highest performance. Based on our best model, we classify 825 BCUs as BL Lac candidates and 405 as FSRQ candidates, however, 190 remain without a clear prediction, but the percentage of BCUs in 4FGL is reduced to 5.1 per cent. The γ-ray photon index, synchrotron peak frequency, and high-energy peak frequency of a large sample are used to investigate the relationship between FSRQs and BL Lacs (LBLs, IBLs, and HBLs).

Road vehicles, particularly cars, are one of the primary sources of CO2 emissions in the transport sector. Shifting to unconventional energy sources such as solar and wind power may reduce their carbon footprints considerably. Consequently, using ammonia as a fuel due to its potential benefits, such as its high energy density, being a carbon-free fuel, and its versatility during storage and transportation, has now grabbed the attention of researchers. However, its slow combustion speed, larger combustion chamber requirements, ignition difficulties, and limited combustion stability are still major challenges. Therefore, authors tried to analyze the combustion pressure of ammonia in a constant-volume combustion chamber across different equivalence ratios by adopting a machine learning approach. While conducting the analysis, the experimental values were assessed and subsequently utilized to predict the induced combustion pressure in a constant-volume combustion chamber across various equivalence ratios. In this research, a two-step prediction process was employed. In the initial step, the Random Forest algorithm was applied to assess the combustion pressure. Subsequently, in the second step, artificial neural network machine learning algorithms were employed to pinpoint the most effective algorithm with a lower root-mean-square error and R2. Finally, Linear Regression illustrated the lowest error in both steps with a value of 1.0, followed by Random Forest.

Random forest, extreme gradient boosting and artificial neural network machine learning methods were used to predict the content of myristic acid (C14:0) and conjugated linoleic acid (CLA) based on the milk fatty acids (FA) obtained through Fourier transform infrared spectroscopy. The methods had similar prediction performance for C14:0 (mean average percentage error < 7%). The low prediction accuracy for CLA was probably due to the poor association between the CLA and the input variables. Therefore, machine learning can be used to predict C14:0 and other saturated FA fromed from similar metabolic pathways

Osteoporosis is a prevalent but asymptomatic condition that affects a large population of the elderly, resulting in a high risk of fracture. Several methods have been developed and are available in general hospitals to indirectly assess the bone quality in terms of mineral material level and porosity. In this paper we describe a new method that uses a medical reflex hammer to exert testing stimuli, an electronic stethoscope to acquire impulse responses from tibia, and intelligent signal processing based on artificial neural network machine learning to determine the likelihood of osteoporosis. The proposed method makes decisions from the key components found in the time-frequency domain of impulse responses. Using two common pieces of clinical apparatus, this method might be suitable for the large population screening tests for the early diagnosis of osteoporosis, thus avoiding secondary complications. Following some discussions of the mechanism and procedure, this paper details the techniques of impulse response acquisition using a stethoscope and the subsequent signal processing and statistical machine learning algorithms for decision making. Pilot testing results achieved over 80% in detection sensitivity.

The applications of artificial intelligence in education became a very attractive topic especially during the COVID-19 pandemic due to the high level of uncertainty surrounded the decision making process within the educational institutions. The objective of this study is to create a model that is able to predict the student's score in the SAT test based on the student's performance in the internal assessments of the school and other demographic attributes. The sample includes 37 students of both genders from a private school in the United Arab Emirates (UAE). The findings suggest that it is possible to implement artificial neural networks to estimate the student's performance in the SAT exam based on internal school data. The model accuracy is 87.4 % however, some attributes can be identified as noise data and can be further removed to increase the accuracy. Scholastic Assessment Test Artificial Neural Network Machine learning Students performance.

Major League Baseball (MLB) is the highest level of professional baseball in the world and accounts for some of the most popular international sporting events. Many scholars have conducted research on predicting the outcome of MLB matches. The accuracy in predicting the results of baseball games is low. Therefore, deep learning and machine learning methods were used to build models for predicting the outcomes (win/loss) of MLB matches and investigate the differences between the models in terms of their performance. The match data of 30 teams during the 2019 MLB season with only the starting pitcher or with all pitchers in the pitcher category were collected to compare the prediction accuracy. A one-dimensional convolutional neural network (1DCNN), a traditional machine learning artificial neural network (ANN), and a support vector machine (SVM) were used to predict match outcomes with fivefold cross-validation to evaluate model performance. The highest prediction accuracies were 93.4%, 93.91%, and 93.90% with the 1DCNN, ANN, SVM models, respectively, before feature selection; after feature selection, the highest accuracies obtained were 94.18% and 94.16% with the ANN and SVM models, respectively. The prediction results obtained with the three models were similar, and the prediction accuracies were much higher than those obtained in related studies. Moreover, a 1DCNN was used for the first time for predicting the outcome of MLB matches, and it achieved a prediction accuracy similar to that achieved by machine learning methods.

Lung cancer is a kind of difficult to diagnose and dangerous cancer. It commonly causes death both men and women so fast accurate analysis of nodules is more important for treatment. Various methods have been used for detecting cancer in early stages. In this paper, machine learning methods compared while detect lung cancer nodule. We applied Principal Component Analysis, K-Nearest Neighbors, Support Vector Machines, Naïve Bayes, Decision Trees and Artificial Neural Networks machine learning methods to detect anomaly. We compared all methods both after preprocessing and without preprocessing. The experimental results show that Artificial Neural Networks gives the best result with 82,43% accuracy after image processing and Decision Tree gives the best result with 93,24% accuracy without image processing.

In the published article, there was an error in the Funding statement. [(RSP2025R351)]. The correct Funding statement appears below.[The author(s) declares that financial support was received for the research, authorship, and/or publication of this article. This research was funded by the Researchers Supporting Project Number (RSP2024R249), King Saud University, Riyadh, Saudi Arabia.]The authors apologize for this error and state that this does not change the scientific conclusions of the article in any way. The original article has been updated.In the published article, there was an error.

We present a novel indoor localisation system that used acoustic sensing. We developed the Acoustic Landmark Locator to determine a person's current room location, within a building. Indoor environments tend to have distinct acoustic properties due to physical structure. Hence rooms in a building can have distinctive acoustic signatures. We found that these acoustic signatures can determine the position of a person. We attempted to identify location based on acoustic sensing of the surrounding indoor environment. We developed a mobile phone application that determined a person's location by measuring the acoustic levels of the surrounding environment. We used a machine learning artificial neural network based algorithm to classify the location of the person, within proximity to a landmark or room. We tested the Acoustic Landmark Locator in an indoor environment. Our tests show that the Acoustic Landmark Locator mobile phone app was able to successfully determine the location of the person carrying the mobile phone, in all test areas. It was also found that background noise caused by the presence of people does distort the landmark acoustic profiles but the artificial neural network based classifier was able to reliably determine the person's room location. Further work will involve investigating how other machine learning approaches can be used to better improve position accuracy.

Human-Machine Interface for Artificial Neural Network Based Machine Tool Process Monitoring

We propose a methodology to compensate the effect of mutual coupling in planar antenna array arrangements using a novel electromagnetic machine learning (ML) technique. The electromagnetic behavior of the array elements and their mutual coupling interactions are described via the antenna current Green's function (ACGF). A ML artificial neural network (ANN) procedure is then applied and combined with the ACGF to optimize array ports and mitigate for mutual coupling effects.