ML Algorithm Research Articles

Review Paper on Smart Video Player

Abstract: In this paper, we have proposed a model for a smart video player which would have two main features .One is the simulation using hand gesture technique and second one is the emotion detection of the video using CNN (ML Algorithm).This model gives an accuracy of 90-96% and hence can be studied for further process .This model yields better results than other models in all aspects.

Analysis of Protein Structure for Drug Repurposing Using Computational Intelligence and ML Algorithm

Proteins are fundamental compounds in biological processes during the analysis of drug target indication for drug repurposing. The identification of relevant features is a necessary step in determining protein structure. A classification technique is used to identify the most important features in a dataset, which is why feature selection is so important. For protein structure prediction, recent research has developed a wide range of new methods to improve accuracy. The authors use principal component analysis (PCA) with correlation-matrix-based feature selection to analyse breast cancer data. In this paper, they discussed a therapeutic agent that is used to reduce the dataset by reduction-based algorithm and after that applied reduced dataset labelled as Standard Gold Dataset on machine learning model to analyze drug target indication. They get the higher accuracy of 92.8%, 93.9%, and 95.3%, each of the three datasets with 200, 500, and 1000 features with SVM with RBF kernel function. Also they found the best result, 97.8%, with the same classifier.

Performance Analysis of ML-based Low Complex CFO Estimation for MIMO-OFDMA Uplink Systems

Carrier frequency offset (CFO) estimation in multiuser multiple-input multiple-output (MIMO) orthogonal frequency-division multiplexing (OFDM) system is investigated in this study. MIMO-OFDM is very sensitive to CFOs due to oscillator frequency mismatch and/or Doppler shift. Inaccurate CFO estimation results in intercarrier interference (ICI) through the loss of orthogonality among subcarriers. In this paper, the performance of the ML and APFE algorithm is analyzed for CFO estimation. ML becomes extremely complex due to the multidimensional exhaustive search issue, which is the basic concern in ML estimation. However, making use of the iterative low complex APFE method, here this multidimensional search is replaced with a sequence of mono-dimensional searches. This results in an estimation algorithm of reasonable complexity which is suitable for practical applications. In addition, ML accuracy is compared with the Cramer-Rao bound (CRB).

8P A gradient boosting decision tree (GBDT) approach to identify potential therapeutic targets

8P A gradient boosting decision tree (GBDT) approach to identify potential therapeutic targets

Extracting structural motifs from pair distribution function data of nanostructures using explainable machine learning

Characterization of material structure with X-ray or neutron scattering using e.g. Pair Distribution Function (PDF) analysis most often rely on refining a structure model against an experimental dataset. However, identifying a suitable model is often a bottleneck. Recently, automated approaches have made it possible to test thousands of models for each dataset, but these methods are computationally expensive and analysing the output, i.e. extracting structural information from the resulting fits in a meaningful way, is challenging. Our Machine Learning based Motif Extractor (ML-MotEx) trains an ML algorithm on thousands of fits, and uses SHAP (SHapley Additive exPlanation) values to identify which model features are important for the fit quality. We use the method for 4 different chemical systems, including disordered nanomaterials and clusters. ML-MotEx opens for a type of modelling where each feature in a model is assigned an importance value for the fit quality based on explainable ML.

Traffic light detection and control with connected vehicle for automated car using ML algorithm

The paper centers to plan a self-driving vehicle model with the assistance of Raspberry Pi as a handling chip. The model will perform five undertaking's that remember self-driving for track, identifying traffic signal, rain identification, fog discovery and will help in impediment recognition and crash evasion. The essential information from this present reality to the vehicle is given by camera. It utilizes a HC-SR04 following sensor module that data to the framework through raspberry pi to our framework which is associated with PC with a similar organization. The gathered information is then prepared and examined and pertinent data and subtleties is made back aware of vehicle for proper moves to be initiated. The gamble of human blunders is in this way kept away from securely and keenly as the vehicle can arrive at the given destination. Path Detection, Obstacle Detection, Traffic Light Detection, Rain and Fog are distinguished. The vehicle is halted when impediment and traffic signal red is noticed. As a section one static is planned with Arduino to make Bluetooth (HC-05) communication between the vehicles. This association between two vehicles will keep away from accidents.

Exploiting Machine Learning Models for Identification of Heart Diseases

Cardiovascular illnesses have surpassed disease as the leading cause of death in industrialized, emerging, and underprivileged countries in recent decades. The mortality rate can be lowered through early detection and effective care of cardiac diseases. However, reliable detection of heart disorders in all conditions and serving doctors with 24-hour medical consultations are not possible since they require more intelligence, time, and talent to train the computer for a specific duty. Artificial intelligence and machine learning algorithms may have a substantial impact on the lives of those who are afflicted with chronic diseases. In this work, we used artificial intelligence to create a hybrid model for the prediction of cardiac disease and flags for prevention. The goal of this research was to develop a prediction model that can identify and characterize a patient's illness. We used data of 68975 patients from the University of California, Irvine's repository and trained several algorithms for CVD illness prediction. In our simulation maximum classifiers have achieved the greatest accuracy, which is a huge accomplishment for our research team. Our proposed hybrid ML algorithm framework improved CVD prediction performance when compared to previously offered prediction models. The proposed methods are decision tree and random forest models, which have a 99.98% accuracy for training, followed by support vector machine, which has a 99.30% accuracy followed by other classifiers which yield higher results in comparison. Pre-existing CVD history was considered the most important factor determining the prediction model's accuracy. Our findings reveal that our CVD prediction model based on machine learning techniques developed for health screening datasets is simple to apply and more accurate. Our proposed classifiers have the best accuracy rate.

Consumption Behavior Prediction Based on Multiobjective Evolutionary Algorithm

Consumption behavior prediction reveals customer attributes, personal preferences, and intrinsic laws. Organizations would benefit from knowing further about customer needs and business desires by monitoring client behavior to provide more precise recommendations and boost acquisition rates. The economics of the customer, buyer groupings, and product quality are only a few of the numerous variables that influence customer behavior. The key issue that has to be resolved at this time is how to filter out useful information from these vast amounts of data to forecast customer behavior. For customer consumption behavior prediction and analysis with an advanced quantitative research process, we proposed the multiobjective evolutionary algorithm, which significantly boosts the accuracy of consumption behavior predictions. The dataset is initially gathered based on consumer preferences and behaviors as the essential information for the entire prediction model. Min-max normalization is used as a component of the preprocessing of the data to get the elimination of redundant and superfluous data. The Word2vec model is utilized for feature extraction, and boosted ant colony optimization (BACO) is employed to choose the best features. Utilizing the suggested multiobjective evolutionary algorithm (MOEA), the predictions are made. The suggested system’s performance is assessed, and the metrics are contrasted with more established methods. The findings demonstrate that the suggested MOEA technique performs well than the traditional ML, XGB, AI, and HNB algorithm methods in terms of accuracy (95 percent), quality of prediction (97 percent), precision (99 percent), recall (93 percent), F 1 -score (98 percent), and prediction time (50 seconds). Hence, the outcomes show that the regression model is sustainable. The suggested consumption behavior prediction system has demonstrated its efficiency in boosting profitability.

Machine Learning approach to Predict net radiation over crop surfaces from global solar radiation and canopy temperature data.

As the ground-based instruments for measuring net radiation are costly and need to be handled skillfully, the net radiation data at spatial and temporal scales over Indian subcontinent are scanty. Sometimes, it is necessary to use other meteorological parameters to estimate the value of net radiation, although the prediction may vary based on season, ground cover and estimation method. In this context, artificial intelligence can be used as a powerful tool for predicting the data considering past observed data. This paper proposes a novel method to predict the net radiation for five crop surfaces using global solar radiation and canopy temperature. This contribution includes the generation of real-time data for five crops grown in West Bengal state of India. After manual analysis and data preprocessing, data normalization has been done before applying machine learning approaches for training a robust model. We have presented the comparison in various machine learning algorithm such as ridge and spline regression, random forest, ensemble and deep neural networks. The result shows that the gradient boosting regression and ridge regression are outperforming other ML approaches. The estimated predictors enable to reduce the number of resources in terms of time, cost and manpower for proper net radiation estimation. Thus, the problem of predicting net radiation over various crop surfaces can be sorted out through ML algorithm.

The Role of Artificial Intelligence in Decoding Speech from EEG Signals: A Scoping Review.

Background: Brain traumas, mental disorders, and vocal abuse can result in permanent or temporary speech impairment, significantly impairing one’s quality of life and occasionally resulting in social isolation. Brain–computer interfaces (BCI) can support people who have issues with their speech or who have been paralyzed to communicate with their surroundings via brain signals. Therefore, EEG signal-based BCI has received significant attention in the last two decades for multiple reasons: (i) clinical research has capitulated detailed knowledge of EEG signals, (ii) inexpensive EEG devices, and (iii) its application in medical and social fields. Objective: This study explores the existing literature and summarizes EEG data acquisition, feature extraction, and artificial intelligence (AI) techniques for decoding speech from brain signals. Method: We followed the PRISMA-ScR guidelines to conduct this scoping review. We searched six electronic databases: PubMed, IEEE Xplore, the ACM Digital Library, Scopus, arXiv, and Google Scholar. We carefully selected search terms based on target intervention (i.e., imagined speech and AI) and target data (EEG signals), and some of the search terms were derived from previous reviews. The study selection process was carried out in three phases: study identification, study selection, and data extraction. Two reviewers independently carried out study selection and data extraction. A narrative approach was adopted to synthesize the extracted data. Results: A total of 263 studies were evaluated; however, 34 met the eligibility criteria for inclusion in this review. We found 64-electrode EEG signal devices to be the most widely used in the included studies. The most common signal normalization and feature extractions in the included studies were the bandpass filter and wavelet-based feature extraction. We categorized the studies based on AI techniques, such as machine learning and deep learning. The most prominent ML algorithm was a support vector machine, and the DL algorithm was a convolutional neural network. Conclusions: EEG signal-based BCI is a viable technology that can enable people with severe or temporal voice impairment to communicate to the world directly from their brain. However, the development of BCI technology is still in its infancy.

Molecular dipole moment learning via rotationally equivariant derivative kernels in molecular-orbital-based machine learning.

This study extends the accurate and transferable molecular-orbital-based machine learning (MOB-ML) approach to modeling the contribution of electron correlation to dipole moments at the cost of Hartree-Fock computations. A MOB pairwise decomposition of the correlation part of the dipole moment is applied, and these pair dipole moments could be further regressed as a universal function of MOs. The dipole MOB features consist of the energy MOB features and their responses to electric fields. An interpretable and rotationally equivariant derivative kernel for Gaussian process regression (GPR) is introduced to learn the dipole moment more efficiently. The proposed problem setup, feature design, and ML algorithm are shown to provide highly accurate models for both dipole moments and energies on water and 14 small molecules. To demonstrate the ability of MOB-ML to function as generalized density-matrix functionals for molecular dipole moments and energies of organic molecules, we further apply the proposed MOB-ML approach to train and test the molecules from the QM9 dataset. The application of local scalable GPR with Gaussian mixture model unsupervised clustering GPR scales up MOB-ML to a large-data regime while retaining the prediction accuracy. In addition, compared with the literature results, MOB-ML provides the best test mean absolute errors of 4.21 mD and 0.045 kcal/mol for dipole moment and energy models, respectively, when training on 110 000 QM9 molecules. The excellent transferability of the resulting QM9 models is also illustrated by the accurate predictions for four different series of peptides.

Analysis of Various ML Algorithm with the use of Big Data Analytics: A Review

Recent years have seen a meteoric rise in research attention paid to the big data across all areas of computer science. Methods of instruction in particular communicate vital prospects and possibility of radical change in a number of fields of thought and novel instructional approaches that tackle a wide range of problems. Recent developments in machine learning are discussed in this paper. Learning in the age of big data. New machine concepts are also discussed. Approaches to education, with a concentration on more approaches to learning are presented, such as representation active learning, deep learning, transfer learning, and other types of learning. An overview and discussion of the current state of machine learning tools, accompanying difficulties, and possible solutions are in addition to being shown and contrasted with the viewpoint data mining.

185. A predictive model for C5 palsy after instrumented cervical fusion

<h3>BACKGROUND CONTEXT</h3> C5 palsy is a common postoperative complication of cervical fusion, occurring in 5-24% of patients. Although their prognosis is typically favorable, patients who develop C5 palsy are often unable to perform basic activities of daily living, resulting in increased health care costs and decreased quality of life. Accurate prediction of C5 palsy is thus of clinical utility. Predictive models assessing C5 palsy remain rare in the literature. <h3>PURPOSE</h3> We aim to build an ensemble ML algorithm for prediction of C5 palsy after instrumented cervical fusion. Furthermore, we aim to compare the performance of this model with logistic regression and other standard ML models. We hypothesize that the ML model will identify novel predictive features. <h3>STUDY DESIGN/SETTING</h3> Retrospective, cohort study. <h3>PATIENT SAMPLE</h3> Adult patients who underwent instrumented cervical fusion at a tertiary-care academic medical center between 2013-2020 were included. <h3>OUTCOME MEASURES</h3> The primary outcome was C5 palsy within index admission or follow-up encounters. <h3>METHODS</h3> We developed an ensemble ML model predicting C5 palsy risk using AutoPrognosis, an automated ML framework that configures an ensemble of ML-based models and traditional statistics. We compared this model with logistic regression and four ML models (XGBoost, gradient boosting, AdaBoost, random forest). Discrimination was assessed using area under the receiver operating characteristic curve (AUROC). Calibration was assessed using calibration slope and calibration intercept. We additionally ranked importance of each included feature for model performance utilizing a partial dependence function. <h3>RESULTS</h3> A total of 1,024 patients were included in this study, with 52 cases of C5 palsy (5.1%). The ensemble model demonstrated the highest discrimination (AUROC: 0.733) compared to regression (AUROC: 0.710). This model was well-calibrated with a calibration slope of 0.773 and a calibration intercept of -0.025. The following features were the most important for ensemble model performance: preoperative nonmotor neurologic symptoms, spinal trauma, corpectomy, spinal malignancy, posterior fusion, presence of paresthesias, psychiatric comorbidity, fixation for spinal fracture, degenerative disc disease, interbody fusion. Five of the 10 most important features for the ensemble model were markedly less important for regression. <h3>CONCLUSIONS</h3> We report a well-calibrated ensemble ML model that predicts C5 palsy after instrumented cervical fusion with high discrimination. This represents the first ML algorithm for prediction of C5 palsy after cervical fusion to our knowledge. Individualized risk prediction for patients may facilitate improved preoperative patient counseling and risk stratification. This tool may also aid with addressing potentially modifiable risk factors such as psychiatric comorbidity, helping to decrease likelihood of C5 palsy. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.

Cervical typing in preterm delivery

Preterm delivery remains an enigma and is the leading cause of neonatal morbidity and mortality. To assess the cervix during pregnancy by clinical methods and cervical typing &amp; to determine its use for the prediction of preterm delivery and cervical insufficiency. This was a prospective study conducted on 236 patients attending ANC clinic of the Department of Obstetrics &amp; Gynaecology. Subjects were of different age, parity and gestation period from 16th-36th weeks. History and clinical examination was performed along with cervical typing by doing one PV examination randomly during their visit. Data was tabulated as per cervical typing and followed the patient upto delivery. The data was divided into two groups according to risk associated with the cervical typing (group1: I-V &amp; group 2: VI-IX) and further according to preterm or term delivery. Then both the groups were compared according to cervical typing to know whether low risk or high risk. Statistical tests were applied know whether the difference in obstetrical outcome is statistically significant. Out of 236, 210 patients delivered at term and 26 had preterm delivery. Chi-square Naïve Bayes ML algorithm were applied and showed that this classification according to cervical types is reliable for interpreting the possibility of preterm delivery and is easy to practice even by peripheral doctors at primary health centres in remote areas.

SORG algorithm to predict 3- and 12-month survival in metastatic spinal disease: a cross-sectional population-based retrospective study.

In this study, we wished to compare statistically the novel SORG algorithm in predicting survival in spine metastatic disease versus currently used methods. We recruited 40 patients with spinal metastatic disease who were operated at Geneva University Hospitals by the Neurosurgery or Orthopedic teams between the years of 2015 and 2020. We did an ROC analysis in order to determine the accuracy of the SORG ML algorithm and nomogram versus the Tokuhashi original and revised scores. The analysis of data of our independent cohort shows a clear advantage in terms of predictive ability of the SORG ML algorithm and nomogram in comparison with the Tokuhashi scores. The SORG ML had an AUC of 0.87 for 90days and 0.85 for 1year. The SORG nomogram showed a predictive ability at 90days and 1year with AUCs of 0.87 and 0.76 respectively. These results showed excellent discriminative ability as compared with the Tokuhashi original score which achieved AUCs of 0.70 and 0.69 and the Tokuhashi revised score which had AUCs of 0.65 and 0.71 for 3months and 1year respectively. The predictive ability of the SORG ML algorithm and nomogram was superior to currently used preoperative survival estimation scores for spinal metastatic disease.

Electric Vehicle Accident Alert System

Electric Vehicle Accident Alert System is a Machine Learning Integrated, IoT based real-time alerting system, operating on ESP8266 microcontroller. Our designed system acts as an alerting system in its truest sense. It uses KNN classification algorithm to analyse the GPS location of the vehicle and generate a prediction indicating whether the vehicle is in an accident-prone area or not. And, if the accident occurs, the Electric Vehicle Accident Alert System sends a SOS alert message with a link to google maps (for directions), to the emergency contact saved in the system. The whole system is based on a plug n play concept, i.e., the system would be ready to use once powered up. To provide this agility and flexibility, we have designed a registration web portal for the device which registers the user and their device with a unique UID on the Cloud. After successful registration, the system becomes ready and can be simply fitted inside a vehicle to be used. Moreover,the ML algorithm saves those GPS coordinates in its dataset to further improve the accuracy of the prediction. The whole IoT stack integrated with ML algorithm, Web Portal and a Cloud server (implemented on Google Firebase), makes our project a self-improving, agile and a user-friendly system.

A Study on Machine Learning and Supervised and Deep Learning Algorithms to Predict the Risk of Patients

Technological innovation is adopted sequentially in the medical sciences and health sciences. Due to the innovation of devices, methods, researches, the medical science sector is developing rapidly and its impact helps health professionals to identify diseases easily, predict patients' future diseases probabilities, etc. Using the Framingham dataset, a model built where the ML classification algorithm Linear regression, Logistic regression, SVC, Decision tree, Random Forest, Naive Bayes algorithms, have been used to predict the possibilities of a patient’s next ten-year coronary heart disease risk. DL model Artificial neural networks and the robust ML algorithm impact learning are also used in this research to find the best model and comparison between ML and DL models. After accessing all the ML models, the Logistic regression was found the best effective one with an accuracy score of 0.85063. The Artificial Neural networks and the impact learning provide an accuracy score of 0.84061 and 0.84971 respectively. The aim is to find out the best model which can be easily adopted.

Realization of English Instructional Resources Clusters Reconstruction System Using the Machine Learning Model.

Based on ML algorithm, this paper puts forward a method that can search instructional resources through keyword indexing technology, and then cluster and recombine the related results and present them centrally. In this paper, the semantic processing of user query based on the subject index of educational resources is adopted to make up for the deficiency of query semantics, solve the problem of mismatch between query words and document words, and improve the recall and precision of resource retrieval. It is proposed to select the category feature items manually and establish the category feature model. In the environment of small sample set, the weight of category feature items is trained by ML method. The research shows that the user rating of this system is ideal, reaching 93.21% at the highest. In addition, the stability of the system can still reach 89.31% under the condition of relatively large usage, and its performance is excellent. This system can effectively solve the problem of scattered distribution of English instructional resources and make the presentation of knowledge more in line with the needs of users, thereby further improving the utilization rate of English instructional resources and users' satisfaction.

Rapid Quantification of SARS-Cov-2 Spike Protein Enhanced with a Machine Learning Technique Integrated in a Smart and Portable Immunosensor.

An IoT-WiFi smart and portable electrochemical immunosensor for the quantification of SARS-CoV-2 spike protein was developed with integrated machine learning features. The immunoenzymatic sensor is based on the immobilization of monoclonal antibodies directed at the SARS-CoV-2 S1 subunit on Screen-Printed Electrodes functionalized with gold nanoparticles. The analytical protocol involves a single-step sample incubation. Immunosensor performance was validated in a viral transfer medium which is commonly used for the desorption of nasopharyngeal swabs. Remarkable specificity of the response was demonstrated by testing H1N1 Hemagglutinin from swine-origin influenza A virus and Spike Protein S1 from Middle East respiratory syndrome coronavirus. Machine learning was successfully used for data processing and analysis. Different support vector machine classifiers were evaluated, proving that algorithms affect the classifier accuracy. The test accuracy of the best classification model in terms of true positive/true negative sample classification was 97.3%. In addition, the ML algorithm can be easily integrated into cloud-based portable Wi-Fi devices. Finally, the immunosensor was successfully tested using a third generation replicating incompetent lentiviral vector pseudotyped with SARS-CoV-2 spike glycoprotein, thus proving the applicability of the immunosensor to whole virus detection.

Preprocedural determination of an occlusion pathomechanism in endovascular treatment of acute stroke: a machine learning-based decision

ObjectiveTo evaluate whether an occlusion pathomechanism can be accurately determined by common preprocedural findings through a machine learning-based prediction model (ML-PM).MethodsA total of 476 patients with acute stroke who underwent...