Background

Asthma exacerbations in children pose a significant burden on healthcare systems and families. While traditional risk assessment tools exist, artificial intelligence (AI) offers the potential for enhanced prediction models.

Published at LXer: Reor is a private AI personal knowledge management tool. Think of it as a notes program on steroids. Each note is saved as a Markdown file to a �vault� directory on your machine....

This Wednesday I’ll be speaking at AltConf on Machine Learning on iOS: Integrating IBM Watson with Core ML. Here’s the abstract I submitted: Apple recently announced a partnership with IBM to integrate Core ML with Watson, allowing visual recognition to run locally on iOS devices. The ability to use machine learning while offline opens up […]

The post Altconf 2018 – Machine Learning on iOS: Integrating IBM Watson with Core ML first appeared on Tom Markiewicz.

Earlier this year I attended AltConf in San Jose, a community-driven and supported event held alongside Apple’s WWDC. IBM sponsored the event and offered numerous workshops to attendees. In addition to assisting with the workshops and manning the booth, I had a talk accepted on Machine Learning on iOS: Integrating IBM Watson with Core ML. […]

The post Video: Machine Learning on iOS, Integrating IBM Watson with Core ML at AltConf first appeared on Tom Markiewicz.

Fully updated to reflect modern developments in the field, the Fifth Edition of An Introduction to Optimization fills the need for an accessible, yet rigorous, introduction to optimization theory and methods, featuring innovative coverage and a straightforward approach. The book begins with a review of basic definitions and notations while also providing the related fundamental background of linear algebra, geometry, and calculus. With this foundation, the authors explore the essential topics of unconstrained optimization problems, linear programming problems, and nonlinear constrained optimization. In addition, the book includes an introduction to artificial neural networks, convex optimization, multi-objective optimization, and applications of optimization in… [Continue Reading]

The confluence of the internet of things (IoT) with smart grids has ushered in a paradigm shift in energy management, promising unparalleled efficiency, economic robustness and unwavering reliability. However, this integrative evolution has concurrently amplified the grid's susceptibility to cyber intrusions, casting shadows on its foundational security and structural integrity. Machine learning (ML) and deep learning (DL) emerge as beacons in this landscape, offering robust methodologies to navigate the intricate cybersecurity labyrinth of IoT-infused smart grids. While ML excels at sifting through voluminous data to identify and classify looming threats, DL delves deeper, crafting sophisticated models equipped to counteract avant-garde cyber offensives. Both of these techniques are united in their objective of leveraging intricate data patterns to provide real-time, actionable security intelligence. Yet, despite the revolutionary potential of ML and DL, the battle against the ceaselessly morphing cyber threat landscape is relentless. The pursuit of an impervious smart grid continues to be a collective odyssey. In this review, we embark on a scholarly exploration of ML and DL's indispensable contributions to enhancing cybersecurity in IoT-centric smart grids. We meticulously dissect predominant cyber threats, critically assess extant security paradigms, and spotlight research frontiers yearning for deeper inquiry and innovation.

To achieve psychological and emotional analysis of massive internet chats, researchers have used statistical methods, machine learning, and neural networks to analyse the dynamic tendencies of texts dynamically. For long readers, the author first compares and explores the differences between the two psychoanalysis algorithms based on the emotion dictionary and machine learning for simple sentences, then studies the expansion algorithm of the emotion dictionary, and finally proposes an extended text psychoanalysis algorithm based on conditional random field. According to the experimental results, the mental dictionary's accuracy, recall, and F-score based on the cognitive understanding of each additional ten words were calculated. The optimisation decreased, and the memory and F-score improved. An <i>F</i>-value greater than 1, which is the most effective indicator for evaluating the effectiveness of a mental analysis problem, can better demonstrate that the algorithm is adaptive in the literature dictionary. It has been proven that this scheme can achieve good results in analysing emotional tendencies and has higher efficiency than ordinary weight-based psychological sentiment analysis algorithms.

Aim/Purpose: This paper presents a machine learning approach for analyzing Course Learning Outcomes (CLOs). The aim of this study is to find a model that can check whether a CLO is well written or not. Background: The use of machine learning algorithms has been, since many years, a prominent solution to predict learner performance in Outcome Based Education. However, the CLOs definition is still presenting a big handicap for faculties. There is a lack of supported tools and models that permit to predict whether a CLO is well written or not. Consequently, educators need an expert in quality and education to validate the outcomes of their courses. Methodology: A novel method named CLOCML (Course Learning Outcome Classification using Machine Learning) is proposed in this paper to develop predictive models for CLOs paraphrasing. A new dataset entitled CLOC (Course Learning Outcomes Classes) for that purpose has been collected and then undergone a pre-processing phase. We compared the performance of 4 models for predicting a CLO classification. Those models are Support Vector Machine (SVM), Random Forest, Naive Bayes and XGBoost. Contribution: The application of CLOCML may help faculties to make well-defined CLOs and then correct CLOs' measures in order to improve the quality of education addressed to their students. Findings: The best classification model was SVM. It was able to detect the CLO class with an accuracy of 83%. Recommendations for Practitioners: We would recommend both faculties’ members and quality reviewers to make an informed decision about the nature of a given course outcome. Recommendation for Researchers: We would highly endorse that the researchers apply more machine learning models for CLOs of various disciplines and compare between them. We would also recommend that future studies investigate on the importance of the definition of CLOs and its impact on the credibility of Key Performance Indicators (KPIs) values during accreditation process. Impact on Society: The findings of this study confirm the results of several other researchers who use machine learning in outcome-based education. The definition of right CLOs will help the student to get an idea about the performances that will be measured at the end of a course. Moreover, each faculty can take appropriate actions and suggest suitable recommendations after right performance measures in order to improve the quality of his course. Future Research: Future research can be improved by using a larger dataset. It could also be improved with deep learning models to reach more accurate results. Indeed, a strategy for checking CLOs overlaps could be integrated.

Aim/Purpose: The study focused on learner sentiments and experiences after using the Moodle assessment module and trained a machine learning classifier for future sentiment predictions. Background: Learner assessment is one of the standard methods instructors use to measure students’ performance and ascertain successful teaching objectives. In pedagogical design, assessment planning is vital in lesson content planning to the extent that curriculum designers and instructors primarily think like assessors. Assessment aids students in redefining their understanding of a subject and serves as the basis for more profound research in that particular subject. Positive results from an evaluation also motivate learners and provide employment directions to the students. Assessment results guide not just the students but also the instructor. Methodology: A modified methodology was used for carrying out the study. The revised methodology is divided into two major parts: the text-processing phase and the classification model phase. The text-processing phase consists of stages including cleaning, tokenization, and stop words removal, while the classification model phase consists of dataset training using a sentiment analyser, a polarity classification model and a prediction validation model. The text-processing phase of the referenced methodology did not utilise tokenization and stop words. In addition, the classification model did not include a sentiment analyser. Contribution: The reviewed literature reveals two major omissions: sentiment responses on using the Moodle for online assessment, particularly in developing countries with unstable internet connectivity, have not been investigated, and variations of the k-fold cross-validation technique in detecting overfitting and developing a reliable classifier have been largely neglected. In this study we built a Sentiment Analyser for Learner Emotion Management using the Moodle for assessment with data collected from a Ghanaian tertiary institution and developed a classification model for future sentiment predictions by evaluating the 10-fold and the 5-fold techniques on prediction accuracy. Findings: After training and testing, the RF algorithm emerged as the best classifier using the 5-fold cross-validation technique with an accuracy of 64.9%. Recommendations for Practitioners: Instead of a closed-ended questionnaire for learner feedback assessment, the open-ended mechanism should be utilised since learners can freely express their emotions devoid of restrictions. Recommendation for Researchers: Feature selection for sentiment analysis does not always improve the overall accuracy for the classification model. The traditional machine learning algorithms should always be compared to either the ensemble or the deep learning algorithms Impact on Society: Understanding learners’ emotions without restriction is important in the educational process. The pedagogical implementation of lessons and assessment should focus on machine learning integration Future Research: To compare ensemble and deep learning algorithms

Currently, Android is a booming technology that has occupied the major parts of the market share. However, as Android is an open-source operating system there are possibilities of attacks on the users, there are various types of attacks but one of the most common attacks found was malware. Malware with machine learning (ML) techniques has proven as an impressive result and a useful method for malware detection. Here in this paper, we have focused on the analysis of malware attacks by collecting the dataset for the various types of malware and we trained the model with multiple ML and deep learning (DL) algorithms. We have gathered all the previous knowledge related to malware with its limitations. The machine learning algorithms were having various accuracy levels and the maximum accuracy observed is 99.68%. It also shows which type of algorithm is preferred depending on the dataset. The knowledge from this paper may also guide and act as a reference for future research related to malware detection. We intend to make use of Static Android Activity to analyse malware to mitigate security risks.

Aim/Purpose: Although cassava is one of the crops that can be grown during the dry season in Northeastern Thailand, most farmers in the region do not know whether the crop can grow in their specific areas because the available agriculture planning guideline provides only a generic list of dry-season crops that can be grown in the whole region. The purpose of this research is to develop a predictive model that can be used to predict suitable areas for growing cassava in Northeastern Thailand during the dry season. Background: This paper develops a decision support system that can be used by farmers to assist them determine if cassava can be successfully grown in their specific areas. Methodology: This study uses satellite imagery and data on land characteristics to develop a machine learning model for predicting suitable areas for growing cassava in Thailand’s Nakhon-Phanom province. Contribution: This research contributes to the body of knowledge by developing a novel model for predicting suitable areas for growing cassava. Findings: This study identified elevation and Ferric Acrisols (Af) soil as the two most important features for predicting the best-suited areas for growing cassava in Nakhon-Phanom province, Thailand. The two-class boosted decision tree algorithm performs best when compared with other algorithms. The model achieved an accuracy of .886, and .746 F1-score. Recommendations for Practitioners: Farmers and agricultural extension agents will use the decision support system developed in this study to identify specific areas that are suitable for growing cassava in Nakhon-Phanom province, Thailand Recommendation for Researchers: To improve the predictive accuracy of the model developed in this study, more land and crop characteristics data should be incorporated during model development. The ground truth data for areas growing cassava should also be collected for a longer period to provide a more accurate sample of the areas that are suitable for cassava growing. Impact on Society: The use of machine learning for the development of new farming systems will enable farmers to produce more food throughout the year to feed the world’s growing population. Future Research: Further studies should be carried out to map other suitable areas for growing dry-season crops and to develop decision support systems for those crops.

Aim/Purpose: In the United States, the Centers for Disease Control and Prevention (CDC) tracks the disease activity using data collected from medical practice's on a weekly basis. Collection of data by CDC from medical practices on a weekly basis leads to a lag time of approximately 2 weeks before any viable action can be planned. The 2-week delay problem was addressed in the study by creating machine learning models to predict flu outbreak. Background: The 2-week delay problem was addressed in the study by correlation of the flu trends identified from Twitter data and official flu data from the Centers for Disease Control and Prevention (CDC) in combination with creating a machine learning model using both data sources to predict flu outbreak. Methodology: A quantitative correlational study was performed using a quasi-experimental design. Flu trends from the CDC portal and tweets with mention of flu and influenza from the state of Georgia were used over a period of 22 weeks from December 29, 2019 to May 30, 2020 for this study. Contribution: This research contributed to the body of knowledge by using a simple bag-of-word method for sentiment analysis followed by the combination of CDC and Twitter data to generate a flu prediction model with higher accuracy than using CDC data only. Findings: The study found that (a) there is no correlation between official flu data from CDC and tweets with mention of flu and (b) there is an improvement in the performance of a flu forecasting model based on a machine learning algorithm using both official flu data from CDC and tweets with mention of flu. Recommendations for Practitioners: In this study, it was found that there was no correlation between the official flu data from the CDC and the count of tweets with mention of flu, which is why tweets alone should be used with caution to predict a flu out-break. Based on the findings of this study, social media data can be used as an additional variable to improve the accuracy of flu prediction models. It is also found that fourth order polynomial and support vector regression models offered the best accuracy of flu prediction models. Recommendations for Researchers: Open-source data, such as Twitter feed, can be mined for useful intelligence benefiting society. Machine learning-based prediction models can be improved by adding open-source data to the primary data set. Impact on Society: Key implication of this study for practitioners in the field were to use social media postings to identify neighborhoods and geographic locations affected by seasonal outbreak, such as influenza, which would help reduce the spread of the disease and ultimately lead to containment. Based on the findings of this study, social media data will help health authorities in detecting seasonal outbreaks earlier than just using official CDC channels of disease and illness reporting from physicians and labs thus, empowering health officials to plan their responses swiftly and allocate their resources optimally for the most affected areas. Future Research: A future researcher could use more complex deep learning algorithms, such as Artificial Neural Networks and Recurrent Neural Networks, to evaluate the accuracy of flu outbreak prediction models as compared to the regression models used in this study. A future researcher could apply other sentiment analysis techniques, such as natural language processing and deep learning techniques, to identify context-sensitive emotion, concept extraction, and sarcasm detection for the identification of self-reporting flu tweets. A future researcher could expand the scope by continuously collecting tweets on a public cloud and applying big data applications, such as Hadoop and MapReduce, to perform predictions using several months of historical data or even years for a larger geographical area.

Aim/Purpose: The aim of this study was to analyze various performance metrics and approaches to their classification. The main goal of the study was to develop a new typology that will help to advance knowledge of metrics and facilitate their use in machine learning regression algorithms Background: Performance metrics (error measures) are vital components of the evaluation frameworks in various fields. A performance metric can be defined as a logical and mathematical construct designed to measure how close are the actual results from what has been expected or predicted. A vast variety of performance metrics have been described in academic literature. The most commonly mentioned metrics in research studies are Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), etc. Knowledge about metrics properties needs to be systematized to simplify the design and use of the metrics. Methodology: A qualitative study was conducted to achieve the objectives of identifying related peer-reviewed research studies, literature reviews, critical thinking and inductive reasoning. Contribution: The main contribution of this paper is in ordering knowledge of performance metrics and enhancing understanding of their structure and properties by proposing a new typology, generic primary metrics mathematical formula and a visualization chart Findings: Based on the analysis of the structure of numerous performance metrics, we proposed a framework of metrics which includes four (4) categories: primary metrics, extended metrics, composite metrics, and hybrid sets of metrics. The paper identified three (3) key components (dimensions) that determine the structure and properties of primary metrics: method of determining point distance, method of normalization, method of aggregation of point distances over a data set. For each component, implementation options have been identified. The suggested new typology has been shown to cover a total of over 40 commonly used primary metrics Recommendations for Practitioners: Presented findings can be used to facilitate teaching performance metrics to university students and expedite metrics selection and implementation processes for practitioners Recommendation for Researchers: By using the proposed typology, researchers can streamline development of new metrics with predetermined properties Impact on Society: The outcomes of this study could be used for improving evaluation results in machine learning regression, forecasting and prognostics with direct or indirect positive impacts on innovation and productivity in a societal sense Future Research: Future research is needed to examine the properties of the extended metrics, composite metrics, and hybrid sets of metrics. Empirical study of the metrics is needed using R Studio or Azure Machine Learning Studio, to find associations between the properties of primary metrics and their “numerical” behavior in a wide spectrum of data characteristics and business or research requirements

Aim/Purpose: To predict the change-proneness of software from the continuous evolution using machine learning methods. To identify when software changes become statistically significant and how metrics change. Background: Software evolution is the most time-consuming activity after a software release. Understanding evolution patterns aids in understanding post-release software activities. Many methodologies have been proposed to comprehend software evolution and growth. As a result, change prediction is critical for future software maintenance. Methodology: I propose using machine learning methods to predict change-prone classes. Classes that are expected to change in future releases were defined as change-prone. The previous release was only considered by the researchers to define change-proneness. In this study, I use the evolution of software to redefine change-proneness. Many snapshots of software were studied to determine when changes became statistically significant, and snapshots were taken biweekly. The research was validated by looking at the evolution of five large open-source systems. Contribution: In this study, I use the evolution of software to redefine change-proneness. The research was validated by looking at the evolution of five large open-source systems. Findings: Software metrics can measure the significance of evolution in software. In addition, metric values change within different periods and the significance of change should be considered for each metric separately. For five classifiers, change-proneness prediction models were trained on one snapshot and tested on the next. In most snapshots, the prediction performance was excellent. For example, for Eclipse, the F-measure values were between 80 and 94. For other systems, the F-measure values were higher than 75 for most snapshots. Recommendations for Practitioners: Software change happens frequently in the evolution of software; however, the significance of change happens over a considerable length of time and this time should be considered when evaluating the quality of software. Recommendation for Researchers: Researchers should consider the significance of change when studying software evolution. Software changes should be taken from different perspectives besides the size or length of the code. Impact on Society: Software quality management is affected by the continuous evolution of projects. Knowing the appropriate time for software maintenance reduces the costs and impacts of software changes. Future Research: Studying the significance of software evolution for software refactoring helps improve the internal quality of software code.

Aim/Purpose: Previous research has generally concentrated on identifying the variables that most significantly influence customer churn or has used customer segmentation to identify a subset of potential consumers, excluding its effects on forecast accuracy. Consequently, there are two primary research goals in this work. The initial goal was to examine the impact of customer segmentation on the accuracy of customer churn prediction in the banking sector using machine learning models. The second objective is to experiment, contrast, and assess which machine learning approaches are most effective in predicting customer churn. Background: This paper reviews the theoretical basis of customer churn, and customer segmentation, and suggests using supervised machine-learning techniques for customer attrition prediction. Methodology: In this study, we use different machine learning models such as k-means clustering to segment customers, k-nearest neighbors, logistic regression, decision tree, random forest, and support vector machine to apply to the dataset to predict customer churn. Contribution: The results demonstrate that the dataset performs well with the random forest model, with an accuracy of about 97%, and that, following customer segmentation, the mean accuracy of each model performed well, with logistic regression having the lowest accuracy (87.27%) and random forest having the best (97.25%). Findings: Customer segmentation does not have much impact on the precision of predictions. It is dependent on the dataset and the models we choose. Recommendations for Practitioners: The practitioners can apply the proposed solutions to build a predictive system or apply them in other fields such as education, tourism, marketing, and human resources. Recommendation for Researchers: The research paradigm is also applicable in other areas such as artificial intelligence, machine learning, and churn prediction. Impact on Society: Customer churn will cause the value flowing from customers to enterprises to decrease. If customer churn continues to occur, the enterprise will gradually lose its competitive advantage. Future Research: Build a real-time or near real-time application to provide close information to make good decisions. Furthermore, handle the imbalanced data using new techniques.

Aim/Purpose: While existing literature has extensively explored factors influencing the success of big data projects and proposed big data maturity models, no study has harnessed machine learning to predict project success and identify the critical features contributing significantly to that success. The purpose of this paper is to offer fresh insights into the realm of big data projects by leveraging machine-learning algorithms. Background: Previously, we introduced the Global Big Data Maturity Model (GBDMM), which encompassed various domains inspired by the success factors of big data projects. In this paper, we transformed these maturity domains into a survey and collected feedback from 90 big data experts across the Middle East, Gulf, Africa, and Turkey regions regarding their own projects. This approach aims to gather firsthand insights from practitioners and experts in the field. Methodology: To analyze the feedback obtained from the survey, we applied several algorithms suitable for small datasets and categorical features. Our approach included cross-validation and feature selection techniques to mitigate overfitting and enhance model performance. Notably, the best-performing algorithms in our study were the Decision Tree (achieving an F1 score of 67%) and the Cat Boost classifier (also achieving an F1 score of 67%). Contribution: This research makes a significant contribution to the field of big data projects. By utilizing machine-learning techniques, we predict the success or failure of such projects and identify the key features that significantly contribute to their success. This provides companies with a valuable model for predicting their own big data project outcomes. Findings: Our analysis revealed that the domains of strategy and data have the most influential impact on the success of big data projects. Therefore, companies should prioritize these domains when undertaking such projects. Furthermore, we now have an initial model capable of predicting project success or failure, which can be invaluable for companies. Recommendations for Practitioners: Based on our findings, we recommend that practitioners concentrate on developing robust strategies and prioritize data management to enhance the outcomes of their big data projects. Additionally, practitioners can leverage machine-learning techniques to predict the success rate of these projects. Recommendation for Researchers: For further research in this field, we suggest exploring additional algorithms and techniques and refining existing models to enhance the accuracy and reliability of predicting the success of big data projects. Researchers may also investigate further into the interplay between strategy, data, and the success of such projects. Impact on Society: By improving the success rate of big data projects, our findings enable organizations to create more efficient and impactful data-driven solutions across various sectors. This, in turn, facilitates informed decision-making, effective resource allocation, improved operational efficiency, and overall performance enhancement. Future Research: In the future, gathering additional feedback from a broader range of big data experts will be valuable and help refine the prediction algorithm. Conducting longitudinal studies to analyze the long-term success and outcomes of Big Data projects would be beneficial. Furthermore, exploring the applicability of our model across different regions and industries will provide further insights into the field.

Product innovation has become a crucial part of the sustainability of the coconut agro-industry in Indonesia, covering upstream and downstream sides. To overcome this challenge, it is necessary to create several model stages using a hybrid method that combines machine learning based on multiple criteria decision making and mass balance analysis. The research case study was conducted in Tembilahan district, Riau province, Indonesia, one of the primary coconut producers in Indonesia. The analysis results showed that potential products for domestic customers included coconut milk, coconut cooking oil, coconut chips, coconut jelly, coconut sugar, and virgin coconut oil. Furthermore, considering the experts, the most potential product to be developed was coconut sugar with a weight of 0.26. Prediction of coconut sugar demand reached 13,996,607 tons/year, requiring coconut sap as a raw material up to 97,976,249.

Carlo Lipizzi’s Societal impacts of artificial intelligence and machine learning offers a critical and comprehensive analysis of artificial intelligence (AI) and machine learning’s effects on society. This book provides a balanced perspective, cutting through the

Oxipital AI’s 3D vision and AI offerings aim to be more convenient and effective through a different method of “training” its products.

An internship with Arm gives you exposure to real work and insight into the Arm innovations that shape extraordinary. Students who thrive at Arm take their love of learning beyond their experience of formal education and develop new ideas. This is the energy that interests us.

Internships at Arm will give you the opportunity to put theory into practice through exciting, intellectually challenging, real-world projects that enrich your personal and technical development while enhancing your future career opportunities.

This internship position is within Machine Learning Group in Arm which works on key technologies for the future of computing. Working on the cutting edge of Arm IP, this Group creates technology that powers the next generation of mobile apps, portable devices, home automation, smart cities, self-driving cars, and much more.

When applying, please make sure to include your most up to date academic transcript.

For a sneak peek what it’s like to work in Arm Lund, please have a look at the following video: http://bit.ly/2kxWMXp

The Role

You will work alongside experienced engineers within one of the IP development teams in Arm and be given real project tasks and will be supported by experienced engineers. Examples of previous project tasks are:

• Developing and trialing new processes for use by the design/verification teams.
• Investigating alternative options for existing design or verification implementations.
• Help to develop a hardware platform that can guide out customers to the best solution.
• Implement complex logic using Verilog to bridge a gap in a system.
• Develop bare metal software to exercise design functionality.
• Verify a complex design, from unit to full SoC level.
• Help to take a platform to silicon.

Arm's Machine Learning Group is seeking for a highly motivated and creative Graduate Software Engineer to join the Cambridge-based applied ML team.

From research, to proof-of-concept development, to deployment on ARM IPs, joining this team, would be a phenomenal opportunity to contribute to the full life-cycle of machine learning projects and understand how state-of-the-art machine learning is used to solve real word problems.

Working closely with field experts in a truly multi-discipline environment, you will have the chance to explore existing or build new machine learning techniques, while helping unpick the complex world of use-cases that are applied on high end mobile phones, TVs, and laptops.

About the role

Your role would be to understand, develope and implement these use case, collaborating with Arm's system architects, and working with our marketing groups to ensure multiple Arm products are molded to work well for machine learning. Also, experience deploying inference in a mobile or embedded environment would be ideal. Knowledge of the theory and concepts involved in ML is also needed, so fair comparisons of different approaches can be made.

As an in depth technical role, you will need to understand the complex applications you analyse in detail and communicate them in their simplest form to help include them in product designs, where you will be able to influence both IP and system architecture.

The post Machine Learning Technology in Predicting Relapse and Implementing Peer Recovery Intervention Before Drug Use Occurs was curated by information for practice.

Small-angle X-ray scattering (SAXS) is widely used to analyze the shape and size of nanoparticles in solution. A multitude of models, describing the SAXS intensity resulting from nanoparticles of various shapes, have been developed by the scientific community and are used for data analysis. Choosing the optimal model is a crucial step in data analysis, which can be difficult and time-consuming, especially for non-expert users. An algorithm is proposed, based on machine learning, representation learning and SAXS-specific preprocessing methods, which instantly selects the nanoparticle model best suited to describe SAXS data. The different algorithms compared are trained and evaluated on a simulated database. This database includes 75 000 scattering spectra from nine nanoparticle models, and realistically simulates two distinct device configurations. It will be made freely available to serve as a basis of comparison for future work. Deploying a universal solution for automatic nanoparticle model selection is a challenge made more difficult by the diversity of SAXS instruments and their flexible settings. The poor transferability of classification rules learned on one device configuration to another is highlighted. It is shown that training on several device configurations enables the algorithm to be generalized, without degrading performance compared with configuration-specific training. Finally, the classification algorithm is evaluated on a real data set obtained by performing SAXS experiments on nanoparticles for each of the instrumental configurations, which have been characterized by transmission electron microscopy. This data set, although very limited, allows estimation of the transferability of the classification rules learned on simulated data to real data.

New software capabilities in RMCProfile allow researchers to study the structure of materials by combining machine learning interatomic potentials and reverse Monte Carlo.

Structure refinement with reverse Monte Carlo (RMC) is a powerful tool for interpreting experimental diffraction data. To ensure that the under-constrained RMC algorithm yields reasonable results, the hybrid RMC approach applies interatomic potentials to obtain solutions that are both physically sensible and in agreement with experiment. To expand the range of materials that can be studied with hybrid RMC, we have implemented a new interatomic potential constraint in RMCProfile that grants flexibility to apply potentials supported by the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) molecular dynamics code. This includes machine learning interatomic potentials, which provide a pathway to applying hybrid RMC to materials without currently available interatomic potentials. To this end, we present a methodology to use RMC to train machine learning interatomic potentials for hybrid RMC applications.

Small-angle X-ray scattering (SAXS) is a widely used method for nanoparticle characterization. A common approach to analysing nanoparticles in solution by SAXS involves fitting the curve using a parametric model that relates real-space parameters, such as nanoparticle size and electron density, to intensity values in reciprocal space. Selecting the optimal model is a crucial step in terms of analysis quality and can be time-consuming and complex. Several studies have proposed effective methods, based on machine learning, to automate the model selection step. Deploying these methods in software intended for both researchers and industry raises several issues. The diversity of SAXS instrumentation requires assessment of the robustness of these methods on data from various machine configurations, involving significant variations in the q-space ranges and highly variable signal-to-noise ratios (SNR) from one data set to another. In the case of laboratory instrumentation, data acquisition can be time-consuming and there is no universal criterion for defining an optimal acquisition time. This paper presents an approach that revisits the nanoparticle model selection method proposed by Monge et al. [Acta Cryst. (2024), A80, 202–212], evaluating and enhancing its robustness on data from device configurations not seen during training, by expanding the data set used for training. The influence of SNR on predictor robustness is then assessed, improved, and used to propose a stopping criterion for optimizing the trade-off between exposure time and data quality.

During beam time at a research facility, alignment and optimization of instrumentation, such as spectrometers, is a time-intensive task and often needs to be performed multiple times throughout the operation of an experiment. Despite the motorization of individual components, automated alignment solutions are not always available. In this study, a novel approach that combines optimisers with neural network surrogate models to significantly reduce the alignment overhead for a mobile soft X-ray spectrometer is proposed. Neural networks were trained exclusively using simulated ray-tracing data, and the disparity between experiment and simulation was obtained through parameter optimization. Real-time validation of this process was performed using experimental data collected at the beamline. The results demonstrate the ability to reduce alignment time from one hour to approximately five minutes. This method can also be generalized beyond spectrometers, for example, towards the alignment of optical elements at beamlines, making it applicable to a broad spectrum of research facilities.

Transition-metal nitro­gen-doped carbons (TM-N-C) are emerging as a highly promising catalyst class for several important electrocatalytic processes, including the electrocatalytic CO2 reduction reaction (CO2RR). The unique local environment around the singly dispersed metal site in TM-N-C catalysts is likely to be responsible for their catalytic properties, which differ significantly from those of bulk or nanostructured catalysts. However, the identification of the actual working structure of the main active units in TM-N-C remains a challenging task due to the fluctional, dynamic nature of these catalysts, and scarcity of experimental techniques that could probe the structure of these materials under realistic working conditions. This issue is addressed in this work and the local atomistic and electronic structure of the metal site in a Co–N–C catalyst for CO2RR is investigated by employing time-resolved operando X-ray absorption spectroscopy (XAS) combined with advanced data analysis techniques. This multi-step approach, based on principal component analysis, spectral decomposition and supervised machine learning methods, allows the contributions of several co-existing species in the working Co–N–C catalysts to be decoupled, and their XAS spectra deciphered, paving the way for understanding the CO2RR mechanisms in the Co–N–C catalysts, and further optimization of this class of electrocatalytic systems.

Characterization of a material structure with pair distribution function (PDF) analysis typically involves refining a structure model against an experimental data set, but finding or constructing a suitable atomic model for PDF modelling can be an extremely labour-intensive task, requiring carefully browsing through large numbers of possible models. Presented here is POMFinder, a machine learning (ML) classifier that rapidly screens a database of structures, here polyoxometallate (POM) clusters, to identify candidate structures for PDF data modelling. The approach is shown to identify suitable POMs from experimental data, including in situ data collected with fast acquisition times. This automated approach has significant potential for identifying suitable models for structure refinement to extract quantitative structural parameters in materials chemistry research. POMFinder is open source and user friendly, making it accessible to those without prior ML knowledge. It is also demonstrated that POMFinder offers a promising modelling framework for combined modelling of multiple scattering techniques.

Data collection at X-ray free electron lasers has particular experimental challenges, such as continuous sample delivery or the use of novel ultrafast high-dynamic-range gain-switching X-ray detectors. This can result in a multitude of data artefacts, which can be detrimental to accurately determining structure-factor amplitudes for serial crystallography or single-particle imaging experiments. Here, a new data-classification tool is reported that offers a variety of machine-learning algorithms to sort data trained either on manual data sorting by the user or by profile fitting the intensity distribution on the detector based on the experiment. This is integrated into an easy-to-use graphical user interface, specifically designed to support the detectors, file formats and software available at most X-ray free electron laser facilities. The highly modular design makes the tool easily expandable to comply with other X-ray sources and detectors, and the supervised learning approach enables even the novice user to sort data containing unwanted artefacts or perform routine data-analysis tasks such as hit finding during an experiment, without needing to write code.

Serial crystallography experiments at synchrotron and X-ray free-electron laser (XFEL) sources are producing crystallographic data sets of ever-increasing volume. While these experiments have large data sets and high-frame-rate detectors (around 3520 frames per second), only a small percentage of the data are useful for downstream analysis. Thus, an efficient and real-time data classification pipeline is essential to differentiate reliably between useful and non-useful images, typically known as `hit' and `miss', respectively, and keep only hit images on disk for further analysis such as peak finding and indexing. While feature-point extraction is a key component of modern approaches to image classification, existing approaches require computationally expensive patch preprocessing to handle perspective distortion. This paper proposes a pipeline to categorize the data, consisting of a real-time feature extraction algorithm called modified and parallelized FAST (MP-FAST), an image descriptor and a machine learning classifier. For parallelizing the primary operations of the proposed pipeline, central processing units, graphics processing units and field-programmable gate arrays are implemented and their performances compared. Finally, MP-FAST-based image classification is evaluated using a multi-layer perceptron on various data sets, including both synthetic and experimental data. This approach demonstrates superior performance compared with other feature extractors and classifiers.

Macromolecular crystallography contributes significantly to understanding diseases and, more importantly, how to treat them by providing atomic resolution 3D structures of proteins. This is achieved by collecting X-ray diffraction images of protein crystals from important biological pathways. Spotfinders are used to detect the presence of crystals with usable data, and the spots from such crystals are the primary data used to solve the relevant structures. Having fast and accurate spot finding is essential, but recent advances in synchrotron beamlines used to generate X-ray diffraction images have brought us to the limits of what the best existing spotfinders can do. This bottleneck must be removed so spotfinder software can keep pace with the X-ray beamline hardware improvements and be able to see the weak or diffuse spots required to solve the most challenging problems encountered when working with diffraction images. In this paper, we first present Bragg Spot Detection (BSD), a large benchmark Bragg spot image dataset that contains 304 images with more than 66 000 spots. We then discuss the open source extensible U-Net-based spotfinder Bragg Spot Finder (BSF), with image pre-processing, a U-Net segmentation backbone, and post-processing that includes artifact removal and watershed segmentation. Finally, we perform experiments on the BSD benchmark and obtain results that are (in terms of accuracy) comparable to or better than those obtained with two popular spotfinder software packages (Dozor and DIALS), demonstrating that this is an appropriate framework to support future extensions and improvements.

High-energy X-ray diffraction methods can non-destructively map the 3D microstructure and associated attributes of metallic polycrystalline engineering materials in their bulk form. These methods are often combined with external stimuli such as thermo-mechanical loading to take snapshots of the evolving microstructure and attributes over time. However, the extreme data volumes and the high costs of traditional data acquisition and reduction approaches pose a barrier to quickly extracting actionable insights and improving the temporal resolution of these snapshots. This article presents a fully automated technique capable of rapidly detecting the onset of plasticity in high-energy X-ray microscopy data. The technique is computationally faster by at least 50 times than the traditional approaches and works for data sets that are up to nine times sparser than a full data set. This new technique leverages self-supervised image representation learning and clustering to transform massive data sets into compact, semantic-rich representations of visually salient characteristics (e.g. peak shapes). These characteristics can rapidly indicate anomalous events, such as changes in diffraction peak shapes. It is anticipated that this technique will provide just-in-time actionable information to drive smarter experiments that effectively deploy multi-modal X-ray diffraction methods spanning many decades of length scales.

One of the biggest challenges in providing relief to people living in poverty is locating them. The availability of accurate and reliable information on the location of impoverished zones is surprisingly lacking for much of the world, particularly on the African continent. Aid groups and other international organizations often fill in the gaps with door-to-door surveys, but these can be expensive and time-consuming to conduct.

read more

The new patented method increases video compression factor by about 3 times, not 40%, against the existing modern standards like H.265, VP9 and AV1, with the same video quality.

Prateek Agarwal serves as a technical lead at Tata Consultancy Services Ltd

Sai Sharanya Nalla recognized as an AI expert with over a decade of experience, including key roles at Nike, AWS and Amex.

Robert Daly serves as a database solutions architect at Amazon Web Services.

Fenix Commerce will expand its AI & ML capabilities to generate even more incremental revenue to its customers

The landscape of AI is rapidly evolving, presenting both opportunities and challenges. From its historical roots to the current AI wars and the pursuit of Artificial General Intelligence (AGI), AI is a force to be reckoned with. Despite remarkable advancements, current AI systems face limitations in adaptive learning and memory, sparking a paradigm shift towards creating more human-like capabilities.

The post Game-Changing Paradigm Shift in Machine Learning! appeared first on WPCult.

BOSTON (October 22, 2024) -- The Free Software Foundation (FSF) has announced today that it is working on a statement of criteria for free machine learning applications, which will require the software, as well as the raw training data and associated scripts, to grant users the four freedoms.

ETSI releases a Technical Report on autonomic network management and control applying machine learning and other AI algorithms

Sophia Antipolis, 5 March 2020

The ETSI Technical Committee on Core Network and Interoperability Testing (TC INT) has just released a Technical Report, ETSI TR 103 626, providing a mapping of architectural components for autonomic networking, cognitive networking and self-management. This architecture will serve the self-managing Future Internet.

The ETSI TR 103 626 provides a mapping of architectural components developed in the European Commission (EC) WiSHFUL and ORCA Projects, using the ETSI Generic Autonomic Networking Architecture (GANA) model.

The objective is to illustrate how the ETSI GANA model specified in the ETSI specification TS 103 195-2 can be implemented when using the components developed in these two projects. The Report also shows how the WiSHFUL architecture augmented with virtualization and hardware acceleration techniques can implement the GANA model. This will guide implementers of autonomics components for autonomic networks in their optimization of their GANA implementations.

The TR addresses autonomic decision-making and associated control-loops in wireless network architectures and their associated management and control architectures. The mapping of the architecture also illustrates how to implement self-management functionality in the GANA model for wireless networks, taking into consideration another Report ETSI TR 103 495, where GANA cognitive algorithms for autonomics, such as machine learning and other AI algorithms, can be applied.

A multi-species benchmark for training and validating mass spectrometry proteomics machine learning models  Nature.com

Sunnyvale, CA – Deep learning will transform medicine, but not in the way that many advocates think. Biological complexity, rare mutations and confounding factors work against us, so that even if we sequence 100,000 genomes, it won’t be enough. Brendan Frey is engineering the future of personalized medicine. A professor in the University of Toronto’s Faculty […]