Improved Machine Learning Approach Research Articles

Automated identification of steel weld defects, a convolutional neural network improved machine learning approach

Automated identification of steel weld defects, a convolutional neural network improved machine learning approach

Validation of an Integrated Genetic-Epigenetic Test for the Assessment of Coronary Heart Disease.

Coronary heart disease (CHD) is the leading cause of death in the world. Unfortunately, many of the key diagnostic tools for CHD are insensitive, invasive, and costly; require significant specialized infrastructure investments; and do not provide information to guide postdiagnosis therapy. In prior work using data from the Framingham Heart Study, we provided in silico evidence that integrated genetic-epigenetic tools may provide a new avenue for assessing CHD. In this communication, we use an improved machine learning approach and data from 2 additional cohorts, totaling 449 cases and 2067 controls, to develop a better model for ascertaining symptomatic CHD. Using the DNA from the 2 new cohorts, we translate and validate the in silico findings into an artificial intelligence-guided, clinically implementable method that uses input from 6 methylation-sensitive digital polymerase chain reaction and 10 genotyping assays. Using this method, the overall average area under the curve, sensitivity, and specificity in the 3 test cohorts is 82%, 79%, and 76%, respectively. Analysis of targeted cytosine-phospho-guanine loci shows that they map to key risk pathways involved in atherosclerosis that suggest specific therapeutic approaches. We conclude that this scalable integrated genetic-epigenetic approach is useful for the diagnosis of symptomatic CHD, performs favorably as compared with many existing methods, and may provide personalized insight to CHD therapy. Furthermore, given the dynamic nature of DNA methylation and the ease of methylation-sensitive digital polymerase chain reaction methodologies, these findings may pave a pathway for precision epigenetic approaches for monitoring CHD treatment response.

An Improved Machine Learning Approach for Optimizing Dust Concentration Estimation in Open-Pit Mines.

Dust is a severe environmental issue in open-pit mines, and accurate estimation of its concentration allows for viable solutions for its control and management. This research proposes a machine learning-based solution for accurately estimating dust concentrations. The proposed approach, tested using real data from the Haerwusu open-pit coal mine in China, is based upon the integrated random forest-Markov chain (RF-MC) model. The random forest method is used for estimation, while the Markov chain is used for estimation correction. The wind speed, temperature, humidity, and atmospheric pressure are used as inputs, while PM2.5, PM10, and TSP are taken as estimated outputs. A detailed procedure for implementing the RF-MC is presented, and the estimated performance is analyzed. The results show that after correction, the root mean squared error significantly decreased from 7.40 to 2.56 μg/m3 for PM2.5, from 15.73 to 5.28 μg/m3 for PM10, and from 18.99 to 6.27 μg/m3 for TSP, and the Pearson correlation coefficient and the mean absolute error also improved considerably. This work provides an improved machine learning approach for dust concentration estimation in open-pit coal mines, with a greater emphasis on simplicity and rapid model updates, which is more applicable to ensure the prudent use of water resources and overall environmental conservation, both of which are advantageous to green mining.

Athlete’s Performance Analysis Based on Improved Machine Learning Approach on Wearable Devices

Today, every nation strives for international recognition in a variety of sports. Governments invest in games and sports to raise the performance of their teams and athletes to get notoriety. Numerous people are involved in sports execution, including team management, coaches, and biomechanists who monitor athlete fitness and work to achieve remarkable results. Performance analysis is greatly aided by technological integration in sports management. The performance analysis of athletes is evaluated in this research using an upgraded machine learning approach on Improved Machine Learning approach on Wearable Devices (IMLA-WD). This design strategy utilizes wearable devices to collect health data, which is then fed into a machine-learning model to monitor athletes' progress. The athletes' performance is evaluated using standard machine learning methods, and the deep neural network monitors their health status. With a health prediction accuracy of 98.65%, the statistical findings of the proposed model demonstrate the highest performance compared to existing methodologies.

Real-Time State of Charge Estimation of Lithium-Ion Batteries Using Optimized Random Forest Regression Algorithm

This paper presents an improved machine learning approach for the accurate and robust state of charge (SOC) in electric vehicle (EV) batteries using differential search optimized random forest regression (RFR) algorithm. The precise SOC estimation confirms the safety and reliability of EV. Nevertheless, SOC is influenced by numerous factors which cannot be measured directly. RFR is suitable for real-time SOC estimation due to its robustness to noise, overfitting issues and capacity to work with huge datasets. However, proper selection of RFR architecture and hyper-parameters combination remains a key issue to be explored. Hence, a differential search algorithm (DSA) is employed to search for the optimal values of trees and leaves in the RFR algorithm. DSA optimized RFR eliminates the utilization of the filter in data pre-processing steps and does not require a detailed understanding and knowledge about battery chemistry, rather only needs sensors to monitor battery voltage and current. The developed approach is validated at room temperature using two types of lithium-ion batteries under a pulse discharge test. In addition, the proposed model is verified under varying temperature settings under EV drive cycles. The experimental results demonstrate that the DSA optimized RFR algorithm achieves RMSE of 0.382% in the HPPC test using LiNMC battery. Besides, the proposed method obtains satisfactory outcomes in EV drive cycles, estimating MAE of 0.193% and 0.346% in DST and FUDS cycles, respectively, at 25°C.

Enhanced Machine Learning Approach for Accurate and Fast Resolution of Inverse Scattering Problem in Breast Cancer Detection

An improved machine learning approach is presented in this paper to guarantee the fast convergence of the Born Iterative Method, even in the presence of strong scatterers, by assuming a single operating frequency and a reduced number of antennas in the scattering setup. The initial estimation of the dielectric profile, provided by the Born Iterative Method, was processed by a specific convolutional neural network to improve the reconstruction using a fast machine learning approach. To ensure rapid convergence, a proper choice of the initial guess in terms of the minimum permittivity value over the entire domain was also made. Numerical validations on realistic breast phantoms were illustrated, demonstrating an average error of 2.4% and an accuracy greater than 96% for all considered tests, even when considering a single operating frequency and a reduced amount of training data.

An improved machine learning approach for predicting granular flows

Granular flow is widely practised in many industry processes. The previous prediction methods of granular flow are limited in efficiency, including conventional Discrete Element Modelling (DEM) where direct computation of particle collisions is very time-consuming; and recent machine learning (ML) approach where particle positions were straight predicted and particle–particle and particle–wall collisions were not considered separately, likely compromising the prediction accuracy. In this study, an improved ML approach is developed for predicting granular flows efficiently in terms of computational speed and accuracy. In the proposed ML approach, inspired by Newton’s second law’s concept – from particle acceleration to calculate velocity and then position, a new continuous convolutional neural network (CNN) is established to predict the particles’ accelerations first based on the particle–particle and particle–wall collisions separately, and the particle accelerations are used for calculating particle velocities and finally particle positions. The ML approach is applied to a typical granular flow - particle packing for demonstration. A dataset of 100 scenes of DEM simulations in one scenario is established for network training and examination. The results show that, in long-sequence predictions, the accuracy of the ML approach is three times higher than the previous ML approach. The effects of hyperparameters in the network are quantified. Then the ML approach with the optimized hyperparameters is used in additional three scenarios for further examining the prediction effectiveness. It is indicated that the improved ML approach can satisfactorily capture the morphology of granular flows under three new different scenarios; and the computational cost is only one-seventh compared to the DEM approach under the present conditions. The ML approach provides a simple and time-effective tool for simulating granular flows.

Abstract LB533: TEtor: Improved machine learning approach unveils the clues of cancer lurking in individual TCR repertoire

Abstract T cell receptor (TCR) repertoires are formed by random recombination of genomic precursor elements, some of which mediate recognition of cancer-associated antigens. Identifying these cancer-associated TCRs (caTCRs) could facilitate the early detection of cancers and trace them back to their origins. Due to the complicated processes of T cell immune response, a practical strategy to obtain caTCRs is developing computational methods to identify caTCRs from population and/or individual TCR repertoires. Several machine learning-based methods have been proposed during the latest years and have promoted a series of interesting findings. However, these methods can be further improved by overcoming two vital weaknesses on the design of machine learning framework. First, to facilitate the computational processes, the existing machine learning frameworks prefer to decompose the original TCR sequences into length-fixed amino acid fragments, but the lengths of cancer-associated motifs vary from 3 to 10 amino acids. Second, the correlations among TCRs in the same repertoire should be considered, which are not supported in the existing machine learning frameworks. Thus, we developed a multi-instance learning framework incorporating a convolutional neural network (CNN) model, named TEtor, to solve the issues above. TEtor improves the predictions of caTCRs in pan-cancer TCR repertoires on two aspects. On one hand, the CNN model is adopted with the alternative convolutional filters and 1-max pooling operations. Such design of CNN model is able to handle the amino acid fragments with different lengths, and thus realizing the identification of cancer-associated motifs with various lengths. On the other hand, the multi-instance learning framework calculates the correlations among TCRs and assigns an adjusted weight for each TCR according to the prediction and training process. When training this machine learning model, to speed up the convergence, we adopted a filtering step, which only retains the top 500 abundant TCR sequences in each repertoire. To validate the performance of TEtor, we conducted a series of experiments on several cohorts of patients encompassing 11 different cancers. Compared to the existing machine learning methods, TEtor achieved better prediction performance on accuracy, sensitivity, specificity and areas under the curves (AUCs) on most of the cohorts. For all of the tested cancer types, TEtor obtained an AUC above 0.7 (e.g. thyroid cancer, 0.97; lung cancer, 0.90; gastrointestinal cancer, 0.9). In conclusion, TEtor is a reasonable and practical framework. It improves the caTCR prediction and could further support the accurate detection of early cancers from individual TCR repertoire data. Citation Format: Ying Xu, Xinyang Qian, Xuanping Zhang, Jiayin Wang. TEtor: Improved machine learning approach unveils the clues of cancer lurking in individual TCR repertoire [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB533.

An Improved Machine Learning Approach for Predicting Granular Flows

An Improved Machine Learning Approach for Predicting Granular Flows

An improved machine learning approach to estimate hemicellulose, cellulose, and lignin in biomass

Biomass consists of predominantly a biochemical mixture of three major lignocellulosic components: hemicellulose, cellulose, and lignin. The amounts of these three major polymeric components in biomass affect the functioning of thermochemical processes and thus affecting the efficiency of the biomass-based energy plants. On the other hand, experimental procedures for the hemicellulose, cellulose, and lignin determination are expensive, time-consuming, and not commercially viable. Recently, researchers are developing computational methods to minimize experimental procedures. Artificial neural network (ANN), which is a subset of machine learning, can develop relationships between input-output for non-linear and complex datasets. In this paper, we reported the first comprehensive study on estimating hemicellulose, cellulose, and lignin fractions of biomass using proximate analysis results on as received basis with a machine learning method. The ANN model performed a satisfactory performance for all the major biochemical components with the determination coefficient R2> 0.96.

Automatic handwritten character recognition of Devanagari language: a hybrid training algorithm for neural network

In the field of image processing and Artificial Intelligence the character recognition and handwritten recognition have been emerging as one of the active and challenging research areas. In recent years, there has been a notable development in the research associated with character recognition in handwritten Devanagari documents. To improve the recognition performance, this paper tactics to develop new handwritten character recognition model using an improved machine learning approach. The proposed character recognition model includes four stages like pre-processing, Segmentation, Feature Extraction, and Classification. Initially, the scanned handwritten document for Devanagari language is subjected to pre-processing, which includes stages like, RGB to gray, thresholding, complement of image, morphological operations, linearization and noise removal using Median Filter. Then, the characters of the pre-processed image are segmented using k-means clustering that is a popular method for cluster analysis. Further, the features like, Kirsch Directional Edge, Freeman chain code and neighborhood distance weight using Delaunay Triangulation are extracted from the segmented characters. Subsequently, the classification of characters is done using Neural Network (NN) where the new training algorithm is used, in which the weights are optimized using a hybrid optimization by combining Lion Optimization Algorithm and Grey Wolf Optimization (GWO). Here, the update procedure of GWO is based on LA algorithm and hence, the proposed algorithm is named as Lion Updated GWO (LU-GWO). To the next of the implementation, a valuable comparative analysis confirms the improved performance of the proposed LU-GWO-NN model over conventional methods in classifying the consonants, numerals, and vowels from different characters.

Explicit Link Discovery Scheme Optimized with Ontology Mapping using Improved Machine Learning Approach

Explicit Link Discovery Scheme Optimized with Ontology Mapping using Improved Machine Learning Approach

Diabetes prediction using an improved machine learning approach

This paper deals with a machine-learning model arising from the healthcare sector, namely diabetes progression. The model is reformulated into a regularized optimization problem. The term of the fidelity is the L1 norm and the optimization space of the minimum is constructed by a reproducing kernel Hilbert space (RKSH). The numerical approximation of the model is realized by the Adam method, which shows its success in the numerical experiments (if compared to the stochastic gradient descent (SGD) algorithm).

Image-driven discriminative and generative machine learning algorithms for establishing microstructure–processing relationships

We investigate the methods of microstructure representation for the purpose of predicting processing condition from microstructure image data. A binary alloy (uranium–molybdenum) that is currently under development as a nuclear fuel was studied for the purpose of developing an improved machine learning approach to image recognition, characterization, and building predictive capabilities linking microstructure to processing conditions. Here, we test different microstructure representations and evaluate model performance based on the F1 score. A F1 score of 95.1% was achieved for distinguishing between micrographs corresponding to ten different thermo-mechanical material processing conditions. We find that our newly developed microstructure representation describes image data well, and the traditional approach of utilizing area fractions of different phases is insufficient for distinguishing between multiple classes using a relatively small, imbalanced original dataset of 272 images. To explore the applicability of generative methods for supplementing such limited datasets, generative adversarial networks were trained to generate artificial microstructure images. Two different generative networks were trained and tested to assess performance. Challenges and best practices associated with applying machine learning to limited microstructure image datasets are also discussed. Our work has implications for quantitative microstructure analysis and development of microstructure–processing relationships in limited datasets typical of metallurgical process design studies.

Abstract 3127: To investigate the role of BRAF V600E co-occurring mutations in Langerhans cell histiocytosis (LCH)

Abstract a. Introduction : Langerhans cell histiocytosis (LCH) is an inflammatory myeloid neoplasia caused by alterations of several genes in the MAPKinase pathway. The cell of origin is a white blood cell known as a dendritic cell priviously called as histiocyte characterized by a unique cytoplasmic organelle, the Birbeck granule, and the expression of CD1a and class II major histocompatibility complex molecules . The mutations cause these dendritic cells to attract other white blood cells and cause a “lesion” in any organ of the body. LCH has multi-orgnan involvement affecting skin bones, lymph nodes, brain etc. Systemic symptoms may include fever, bone pain, weight loss, jaundice, diabetes insipidus. b. Description: LCH is a rare sporadic, non-hereditary and non-malignant disease with unknown etiology characterized with a clonal proliferation of pathologic cells.Recent discovery of recurrent somatic mutations, BRAFV600E in MAPK pathway genes at critical stages of myeloid hematopoietic differentiation in LCH patients supports redefinition of the disease as a myeloproliferative disorder and provides opportunities to develop novel approaches to diagnosis and therapy. In the present proposal, we intend to analyze the BRAFV600E status along with other co-occurring mutations in a cohort of pediatric LCH patients by DNA sequencing in bone marrow samples or Peripheral Blood. Additionally we approach the above unanswered questions by applying an improved machine learning approach for classification and defining an algorithm for disease into different clusters for immediate clinical implication and to address its clinical heterogeneity. c. Results : We screened about 5 LCH patients and performed tageted exome seq to asertain the mutational profile of these patients. Sequencing detected BRAF V600E mutation in 3 of 5 samples for a mutation frequency of 60%. IHC based immunofloroscence shows that the intensity of phospho-MEK and phospho-ERK staining shows no significant alterations samples with mutation in BRAF. We are currently investigating mutations in other members of MAK pathways d. Conclusion : BRAF-V600E mutations were detected in few of our LCH patients. Neverless seeing the disease heterogeneity, we expect additionall mutations. More studies are needed to investigate mutations in the MAPK pathways that could efffect the course of disease management.We further plan to integrate the clinical and genomic findings to design a predictive model for the disease outcome. Citation Format: Subhradip Karmakar, R Dhar, K Purkayastha, R Seth, S Karmakar. To investigate the role of BRAF V600E co-occurring mutations in Langerhans cell histiocytosis (LCH) [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3127.

Improved Machine Learning Approach for Wavefront Sensing.

In the adaptive optics (AO) system, to improve the effectiveness and accuracy of wavefront sensing-less technology, a phase-based sensing approach using machine learning is proposed. In contrast to the traditional gradient-based optimization methods, the model we designed is based on an improved convolutional neural network. Specifically, the deconvolution layer, which reconstructs unknown input by measuring output, is introduced to represent the phase maps of the point spread functions at the in focus and defocus planes. The improved convolutional neural network is utilized to establish the nonlinear mapping between the input point spread functions and the corresponding phase maps of the optical system. Once well trained, the model can directly output the aberration map of the optical system with good precision. Adequate simulations and experiments are introduced to demonstrate the accuracy and real-time performance of the proposed method. The simulations show that even when atmospheric conditions D/r0 = 20, the detection root-mean-square of wavefront error of the proposed method is 0.1307 λ, which has a better accuracy than existing neural networks. When D/r0 = 15 and 10, the root-mean-square error is respectively 0.0909 λ and 0.0718 λ. It has certain applicative value in the case of medium and weak turbulence. The root-mean-square error of experiment results with D/r0 = 20 is 0.1304 λ, proving the correctness of simulations. Moreover, this method only needs 12 ms to accomplish the calculation and it has broad prospects for real-time wavefront sensing.

Predicting peptide presentation by major histocompatibility complex class I: an improved machine learning approach to the immunopeptidome

BackgroundTo further our understanding of immunopeptidomics, improved tools are needed to identify peptides presented by major histocompatibility complex class I (MHC-I). Many existing tools are limited by their reliance upon chemical affinity data, which is less biologically relevant than sampling by mass spectrometry, and other tools are limited by incomplete exploration of machine learning approaches. Herein, we assemble publicly available data describing human peptides discovered by sampling the MHC-I immunopeptidome with mass spectrometry and use this database to train random forest classifiers (ForestMHC) to predict presentation by MHC-I.ResultsAs measured by precision in the top 1% of predictions, our method outperforms NetMHC and NetMHCpan on test sets, and it outperforms both these methods and MixMHCpred on new data from an ovarian carcinoma cell line. We also find that random forest scores correlate monotonically, but not linearly, with known chemical binding affinities, and an information-based analysis of classifier features shows the importance of anchor positions for our classification. The random-forest approach also outperforms a deep neural network and a convolutional neural network trained on identical data. Finally, we use our large database to confirm that gene expression partially determines peptide presentation.ConclusionsForestMHC is a promising method to identify peptides bound by MHC-I. We have demonstrated the utility of random forest-based approaches in predicting peptide presentation by MHC-I, assembled the largest known database of MS binding data, and mined this database to show the effect of gene expression on peptide presentation. ForestMHC has potential applicability to basic immunology, rational vaccine design, and neoantigen binding prediction for cancer immunotherapy. This method is publicly available for applications and further validation.

Improvement of Reflection Detection Success Rate of GNSS RO Measurements Using Artificial Neural Network

Global Navigation Satellite System (GNSS) radio occultation (RO) has been widely used in the prediction of weather, climate, and space weather, particularly in the area of tropospheric analyses. However, one of the issues with GNSS RO measurements is that they are interfered with by the signals reflected from the earth’s surface. Many RO events are subject to such interfered GNSS measurements, which are considerably difficult to extract from the GNSS RO measurements. To precisely identify interfered RO events, an improved machine learning approach—a gradient descent artificial neural network (ANN)-aided radio-holography method—is proposed in this paper. Since this method is more complex than most other machine learning methods, for improving its efficiency through the reduction in computational time for near-real-time applications, a scale factor and a regularization factor are also adjusted in the ANN approach. This approach was validated using Constellation Observing System for Meteorology, Ionosphere, and Climate/FC-3 atmPhs (level 1b) data during the period of day of year 172–202, 2015, and its detection results were compared with the flag data set provided by Radio Occultation Meteorology Satellite Application Facilities for the performance assessment and validation of the new approach. The results were also compared with those of the support vector machine method for improvement assessment. The comparison results showed that the proposed method can considerably improve both the success rate of GNSS RO reflection detection and the computational efficiency.

An Improved Machine Learning Approach to Enhance the Predictive Accuracy for Screening Potential Active USP1/UAF1 Inhibitors

DNA repair mechanism is an important mechanism employed by the cancerous cell to survive the DNA damages induced during uncontrolled proliferation of cell and anti-cancer drug treatments. In this context, the Ubiquitin-Specific Proteases (USP1) in complex with Ubiquitin Associated Factor 1(UAF1) plays a key role in the survival of cancerous cell by DNA repair mechanism. Thus, this put forth USP1/UAF1 complex as a striking anti-cancer target for screening of anti-cancer molecule. The current research is aimed to improve the classification accuracy of the existing bioactivity predictive chemoinformatics model for screening potential active USP1/UAF1 inhibitors from high-throughput screening data. The current study employed feature selection method to extract key molecular descriptors from the publicly available high-throughput screening dataset of small molecules that were used to screen active USP1/UAF1 complex inhibitors. This study proposes an improved predictive machine learning approach using the feature selection technique and two class Linear Discriminant Technique (LDA) algorithm to accurately predict the active novel USP1/UAF1 inhibitor compounds.

Neural network approach to assist markings of syllabic nuclei for multi-language database

The manual marking of syllabic nuclei is not difficult for a trained person. For a large amount of speech data, it is time consuming, and human error can introduce inconsistencies. The problem becomes worse when the person is not familiar with the language. To provide an automatic procedure for reducing the marking time and eliminate human errors, an adhoc algorithm was developed to find syllabic nuclei using a set of acoustic features. The initial error rate is 15%. The overall time spent marking automatically and correcting errors manually, however, was less than hand-marking by itself. An improved machine learning approach using a backward-error-propagation multi-layered perceptron, with the same acoustic features as inputs, was applied to the training speech. The hand-corrected markings was used as the teacher. Testing on the training data showed only 3%–5% misalignment. The time required to verify markings has further been reduced. When the training extended to a database of 43 different languages, the marking errors dropped to 2%, less than the errors in the initial hand-markings. Finally, when the network is used to mark 7 h of a multi-language database, there is no significant increase in marking errors.