Top Research Papers on ML

The two most recent extensions for ACL2(ml) can suggest now families of similar function definitions, in addition to the families ofsimilar theorems, and the lemma generation tool implemented in ACL2 (ml) has been improved with a method to generate preconditions using the guard mechanism of ACL2.

dlib-ml contains an extensible linear algebra toolkit with built in BLAS support, and implementations of algorithms for performing inference in Bayesian networks and kernel-based methods for classification, regression, clustering, anomaly detection, and feature ranking.

Java-ML is a collection of machine learning and data mining algorithms, which aims to be a readily usable and easily extensible API for both software developers and research scientists.

The goal of implementing these AI-based algorithms is to increase the possibility of diagnosing a gastrointestinal disease at early stage or the ability to predict the development of a particular condition in advance.

In order to create a more just technological world, the authors need more diverse voices at the table when they create technology and to do this, conventional solutions like reducing barriers to entry and addressing the "leaky pipeline" issues that make mid-career professionals drop out or stall on their way to the top need to be added.

This paper introduces ML.NET: a recently open-sourced machine learning framework allowing developers to author and deploy in their applications complex ML pipelines composed of data featurizers and state of the art machine learning models.

This paper proposes a holistic suite of strategies designed to enhance the security of ML workflows, including advanced data protection techniques like anonymization and encryption, model security enhancements through adversarial training and hardening, and the fortification of infrastructure security via secure computing environments and continuous monitoring.

Basic algorithms of machine learning technique, e.g. support vector machine (SVM) and neural networks, and how they are applied to lithography optimization problems are introduced.

Flink-ML is Flink’s open-source distributed machine learning library, which has been added to the Flink ecosystem in response to the exponential growth of machine learning applications in recent years, and enhances the capabilities of the Flink framework.

The mathematics that is hidden behind the Artificial Intelligence, Machine Learning and Deep Learning that the authors are using today, which are not the unusual milestones it has done thus far are described.

This work presents a novel, machine-learning based approach to blocking one of the three main ways website visitors are tracked online—canvas fingerprinting, using ML-CB to train classifiers on the more-difficult-to-change, underlying source code generating the images.

We describe machine learning (ML) lineage as a framework to holistically capture and connect the required information about ML model development and operations. ML lineage distinguishes between model and prediction levels, conceptually encompassing separate yet interconnected core domains for the project, experiment, model, and prediction.

The clinical predictors that have been provided through "NN model" and "DT model" have the potential through determining the small datasets within biomedical engineering that further helps medical practitioners or healthcare professionals to decide on the medicine and treatment required for a patient.

It is proved theoretically that such a hierarchical system can accelerate training in distributed environments where intra-node communication is cheaper than inter- node communication and that Snap ML achieves the same test loss an order of magnitude faster than any of the previously reported results.

This paper shows how the multiplication a horizontal vector representing a Sparse Distributed Representation of patterns in sensory data by a vertical vector representing a SDR of patterns in context data followed by a pattern recognition operation on the resulting matrix results in the integration of relevant context and in the output of data containing meaning.

This book takes you from the basics of Reinforcement and Q Learning to building Deep Recurrent Q-Network agents that cooperate or compete in a multi-agent ecosystem.

This review paper helps researchers to familiarize themselves with the newest research on ML applications in medicine, recognize their challenges and limitations in this area, and identify future research directions.

This paper systematically incorporate RT domain knowledge into ML and develops an ML framework tailored to the problem, called PAL, which covers a number of additional task sets, each of which has never been proven schedulable by any existing approaches for gFP.

This work has used different algorithms including Deep Support Vector Machine (DSVM), Boosted Decision Tree (BDT), Averaged Perceptron (AP) and Bayes Point Machine (BPM) to build various models, in an attempt to better predict defaulters, and results show that, of all the four models used, DSVM can best predictdefaulters.

The TAK-ML framework which supports data collection, model building, and model execution/employment in tactical environments, as well as a set of initial applications of this framework are described and evaluated, showing the capabilities available, the ease of use of the system, and initial insights into the efficacy of the resulting models and applications.

The report highlights the key challenges which need to be addressed in order to take full advantage of the benefits ML can offer in the areas of personalized medicine and underscores the need for biosensors which provide better quality and quantity of data on which ML may operate.

-----------------------------------------------------------------------------------------------------------------------------------

The author suggests that cross-functional groups comprised of specialists in Clinical Pharmacology, Bioinformatics, or Biomarker Technology were necessary to fully utilize the potential of AI/ML-enabled Translational and Personalized Medicine.

A practical framework on how to read and evaluate medical ML papers that create a new ML model (or diagnostic test): ML-PICO (Machine Learning, Population, Identification, Crosscheck, Outcomes).

A technique is developed that rewrites the query to capture provenance and computes explanations based on that while training the model, and the preliminary evaluation shows the reasonable computational cost of the algorithm.

The results show that the Linear Support Vector Classification (LinearSVC) algorithm, along with the Term Frequency – Inverse Document Frequency (TF-IDF) vectorizer, is well-suited for this multi-class classification task.

By implementing and analyzing an adversarial attack on image classification models, this work demonstrates their detrimental effects on model performance and robustness and underscores the urgent need for proactive measures to safeguard machine learning systems against security threats.

This paper proposes a novel method by rethinking the method of moments (MoM) solving process into a machine learning training process based on the artificial neural network (ANN), based on which machine learningTraining process becomes conventional linear algebra MoM solving process.

This paper presents an alternative approach leveraging a hierarchical planning to configure machine learning pipelines that are unlimited in length and finds its performance to be competitive with other AutoML tools, including TPOT.

The code of the original author was used as a primary resource of this reproducibility attempt and the model attempted to adapt to a new voice with diverse acoustic conditions using VCTK and LJSpeech datasets.

This work has shown that artificial intelligence can be used to predict host phenotypes based on feature selection informed by taxonomy to establish an association between the microbiome and individual characteristics and this relationship can be a proxy to predict various disease states and improve human health.

This study examines how different teams used their in-depth understanding of wearable technologies to accomplish their goals in this study and shows that machine learning may be used to increase energy efficiency and save a lot of money.

ML-ACE extends the existing client-server paradigm for inference serving by providing admission control, preventing user inference requests from over-saturating system resources, in order to systematically schedule ML inference on resource-constrained edge computing platforms.

A taxonomy of ML on Chain considers existing and future potential methods in this area and groups them based on their design characteristics and compares the different 5 groups of solutions across different settings and three ML model types: logistic regression, k-nearest neighbors and neural networks.

Machine Learning is a method which helps in predicting the future or categorized data to help people in making essential decisions and is usable in many industries like sales and marketing, medical sector, healthcare diagnosis, manufacturing, finance, and several other fields.

This paper presents their distributed machine learning subsystem within the Vertica database, Vertica-ML, which includes machine learning functionalities with SQL API which cover a complete data science workflow as well as model management and a set of experiments to evaluate the performance of the machine learning algorithms implemented on top of it.

The Essential Tools of Scientific Machine Learning (Scientific ML) Christopher Rackauckas Affiliation not available April 17, 2023.

In the research reported in this paper, efforts have been made to define and employ heuristic as well as algorithmic rules to conceptualize numerical data produced by normal and faulty jet and rocket engine behavior examples, employed in developing the machine learning system called MLS.

The design, implementation and evaluation of Explain-ML is presented, an Interactive Machine Learning system for Explainable Machine Learning that follows the principles of Human-Centered Machine Learning (HCML) and the results contribute to the understanding and consolidation of the IML principles, ultimately advancing the knowledge in HCML.

Cardiovascular diseases (CD) are the common cause of death worldwide over in developed as well as underdeveloped and developing countries. Early detection and continuous supervision can reduce the mortality rate. Cardiovascular disease diagnosis and accurate diagnosis to enable early treatment. Some of these techniques do not easily diagnose heart diseases at early stages hence, getting treatment late poses a big risk. The present work attempts to better predict this disease from the chest pain symptom, and classify it by designing an efficient machine learning system based on a dataset with 3...

This survey presents the publicly available LoRaWAN frameworks that could be utilized for dataset collection, discusses the required features for efficient resource management with suggested ML methods, and highlights the existing publicly available datasets.

Interpretable machine learning was found to help in teaching many EFL/ESL topics such as dictation, phonetics instruction, language blogging, and all of which contribute to the facilitation of English language learning and teaching processes.

A statistical analysis of the literatures shows that artificial neural network and genetic algorithm are the two most applied ML techniques and the topics in the optimization of device structures and optimization of fabrication processes are more popular.

This study includes the implementation of ML models on a publicly available dataset that evaluates critical risk factors to facilitate the decision process for loan sanctions based on an individual’s relevant characteristics, with SVM being the most suited-algorithm.

The main contribution of this research paper is to present a complete picture of the end-to-end workflows of a production-ready ML pipeline, which can apply to any production ML project though some workflow or steps may differ due to the domain or use-case's demand.

With the technological advancement, the world is going through, at such a high pace, new challenges arise that might require new approaches. ML is, without a doubt, on the lead of technology, as it is evolving with minimal human intervention. However, new models of ML are crucial to preventing hackers from manipulating data. It is true that with new models, new gates for attacks are opened, which is why ML security must follow the same pace in evolving. In this paper, a brief introduction to ML and ML security will be discussed to give an insight on the topic, before going through a more speci...

A thematic evolution towards big data and digital marketing is observed, and thematic clusters such as “digital marketing”, “interpretation”, “prediction”, and “healthcare” stand out.

It is argued that exposing semantics of machine learning algorithms, models, and experiments through a canonical format may pave the way to better interpretability and to realistically achieve the full interoperability of experiments regardless of platform or adopted workflow solution.

This work presents an effort to model the prediction about mentalization from neural activity using predictive modeling, and shows how this approach can help improve the prognosis of psychological disorders such as autism and schizophrenia.

Protein mass fingerprinting by MALDI-TOF MS in combination with machine learning (PhenoMS-ML) permits the identification of response signatures generated in cell cultures upon exposure to well-characterized drugs.