ML Algorithm Research Articles

Extending the Basic Local Independence Model to Polytomous Data

A probabilistic framework for the polytomous extension of knowledge space theory (KST) is proposed. It consists in a probabilistic model, called polytomous local independence model, that is developed as a generalization of the basic local independence model. The algorithms for computing “maximum likelihood” (ML) and “minimum discrepancy” (MD) estimates of the model parameters have been derived and tested in a simulation study. Results show that the algorithms differ in their capability of recovering the true parameter values. The ML algorithm correctly recovers the true values, regardless of the manipulated variables. This is not totally true for the MD algorithm. Finally, the model has been applied to a real polytomous data set collected in the area of psychological assessment. Results show that it can be successfully applied in practice, paving the way to a number of applications of KST outside the area of knowledge and learning assessment.

Predicting Car Availability in Free Floating Car Sharing Systems: Leveraging Machine Learning in Challenging Contexts

Free-Floating Car Sharing (FFCS) services are currently available in tens of cities and countries spread all over the worlds. Depending on citizens’ habits, service policies, and road conditions, car usage profiles are rather variable and often hardly predictable. Even within the same city, different usage trends emerge in different districts and in various time slots and weekdays. Therefore, modeling car availability in FFCS systems is particularly challenging. For these reasons, the research community has started to investigate the applicability of Machine Learning models to analyze FFCS usage data. This paper addresses the problem of predicting the short-term level of availability of the FFCS service in the short term. Specifically, it investigates the applicability of Machine Learning models to forecast the number of available car within a restricted urban area. It seeks the spatial and temporal contexts in which nonlinear ML models, trained on past usage data, are necessary to accurately predict car availability. Leveraging ML has shown to be particularly effective while considering highly dynamic urban contexts, where FFCS service usage is likely to suddenly and unexpectedly change. To tailor predictive models to the real FFCS data, we study also the influence of ML algorithm, prediction horizon, and characteristics of the neighborhood of the target area. The empirical outcomes allow us to provide system managers with practical guidelines to setup and tune ML models.

Detection Algorithm of Compressed Sensing Signal in Generalized Spatial Modulation System

For the generalized spatial modulation in the underdetermined system with the number of transmitting antennas larger than the number of receiving antennas, the activation of the antenna is small and inaccurate. The traditional MMSE algorithm and the ZF algorithm still perform the pseudo-inverse operation on the entire channel matrix, which results in a large number of redundancy. Although the ML algorithm has the best detection performance, the complexity is difficult to meet the actual requirements. In this paper, for the sparse characteristics of generalized spatial modulation signals, a detection algorithm is improved based on the compressed sensing recovery algorithm SWOMP. The algorithm first selects multiple or one active antenna sequences conforming to the spatial modulation to form an index set according to the situation, and then uses the backtracking principle to select the atomic column according to the threshold, rejects the unreliable sequence, and finally uses the minimum mean square error algorithm to detect the modulation symbol according to the activated antenna index. The pseudo-inverse operation of the entire channel matrix is avoided, and the bit error rate is lower than the ZF algorithm, the MMSE algorithm and the OMP algorithm, and the performance of the proposed algorithm is closer to the ML algorithm, which is a better way that balance between complexity and detection performance.

Pruning SMAC search space based on key hyperparameters

SummaryMachine learning (ML) has been widely applied in many areas in recent decades. However, because of its inherent complexity and characteristics, the efficiency and effectiveness of ML algorithm often to be heavily relies on the technical experts' experience and expertise which play a crucial role to optimize hyperparameters of algorithms. Generally, the procedure tuning the exposed hyperparameters of ML algorithm to achieve better performance is called Hyperparameters Optimization. Traditional hyperparameters optimization methods are manually exhaustive search, which is unavailable for high dimensional search spaces and large datasets. Recent automated sequential model‐based optimization led to substantial improvements for this problem, whose core idea is fitting a regression model to describe the importance and dependence of algorithm's performance on certain given hyperparameter setting. Sequential model‐based algorithm configuration (SMAC) is a the‐state‐of‐art approach, which is specified by four components, Initialize, FitModel, SelectConfigurations, and Intensify. In this article, we propose to add a pruning procedure into SMAC approach, it quantifies the importance of hyperparameters by analyzing the performance of a list of promising configurations and reduces search space by discarding noncritical and bad key hyperparameters. To investigate the impact of pruning for model's performance, we conducted experiments on the configuration space constructed by Auto‐Sklearn and compared the effect of run time and pruning ratio with our algorithm. The experiments results verified that, our method made the configuration selected by SMAC more stable and achieved better performance.

Prospective validation of a machine learning algorithm to predict short-term mortality among outpatients with cancer.

2009 Background: Oncologists accurately identify only 35% of patients with cancer who will die in six months. There is an urgent need for automated, accurate prognostic systems to inform treatment and advance care planning in oncology. We assessed the prospective performance of a previously described ML algorithm (Parikh et al, JAMA Netw Open, 2019) to predict short-term mortality in a cohort of general oncology outpatients. Methods: Our prospective cohort consisted of patients aged ≥18 years who had a medical or gynecologic oncology encounter between March 1 and April 30, 2019 in either a tertiary academic practice or one of twelve community practices within a large academic cancer system. We used a retrospectively validated gradient-boosting ML algorithm, based on 559 structured electronic health record (EHR) variables, to predict 180-day mortality prior to each oncology encounter. For patients with multiple encounters, we selected the last encounter to assess performance. We assessed several performance metrics, including area under the receiver operating curve (AUC), area under the precision-recall curve (AUPRC), scaled Brier score (sBrier; a measure of calibration ranging from 0 [random] to 1 [perfect]), and positive predictive value (PPV). Results: Of 25,537 unique patients, median age was 64.4 (interquartile range 53.3 – 73.0), 76.8% were White, 56.5% were treated at a community center, and 4.1% died within 180 days. The ML algorithm had an AUC of 0.89 (95% confidence interval [CI] 0.88-0.90), AUPRC 0.34, and sBrier 0.29. At a prespecified threshold of 40%, observed 180-day mortality was 44.5% (95% CI 40.7 – 48.4%) in the high-risk group vs. 3.0% (95% CI 2.8% – 3.3%) in the low-risk group. There was an 85-fold difference in mortality (13.6% vs. 0.16%) in the top vs. bottom risk quartiles. The model was well-calibrated for mortality risks ≤40% and slightly under-calibrated for mortality risks > 40%. Performance varied across cancer types in the tertiary hospital but did not vary by race or practice type (Table). Conclusions: In this prospective cohort study among outpatients with cancer, a ML prognostic algorithm based on EHR data had better discrimination and calibration that published cancer-specific models. This is one of the first ML prognostic models to be prospectively validated in oncology. [Table: see text]

Genomic alterations associated with the progression from castration-sensitive to castration-resistant metastatic prostate cancer based on machine learning analysis of cell-free DNA genomic profile.

e17596 Background: Metastatic castration-sensitive prostate cancer (mCSPC) eventually progresses to metastatic castration-resistant prostate cancer (mCRPC), which has few treatment options and carries a poor prognosis. We hypothesize that there are specific genomic alterations (GAs) associated with the progression from mCSPC to mCRPC. Methods: Patients (Pts) with mCSPC and mCRPC undergoing next-generation sequencing of cell-free DNA by a CLIA certified lab (G360, Guardant Health Inc., Redwood City, CA) as a part of routine care were retrospectively identified. Principal components analysis, an unsupervised ML algorithm, was used for data exploration and visualization. A combination of feature selection and supervised machine learning classification algorithms were used to identify genes associated with mCRPC. Gene Ontology enrichment analysis was used to identify pathways enriched for mCRPC-associated GAs. Patterns of mCRPC-associated GAs at a gene- and pathway-level were identified by Bayesian networks fitted using an exact structure learning algorithm. Results: 154 Pts with mCSPC and 187 Pts with mCRPC were included. A set of 16 GAs that robustly distinguished mCRPC from mCSPC (PPV = 94%, specificity = 91%) using supervised machine learning algorithms. These GAs, primarily amplifications, corresponded to AR, MAPK signaling, PI3K signaling, G1/S cell cycle, and receptor tyrosine kinases (RTKs). Positive statistical dependencies were observed between genes in these pathways. At a pathway-level, the presence of G1/S GAs in mCRPC samples increased the likelihood of harboring GAs in RTK, MAPK, and PI3K signaling. Limitations: The retrospective nature of our study means that unknown exposures could act as confounding variables, however this is representative of real-world clinical settings. Although the strength of this study is inclusion of clinically annotated patient samples, the limitation is that patients with mCSPC and mCRPC were unmatched. Conclusions: These results provide evidence that progression from mCSPC to mCRPC is associated with stereotyped concomitant gain-of-function in the RTK, PI3K, MAPK, and G1/S pathways in addition to AR. Upon external validation, these hypothesis generating data may warrant further investigation into combinatorial therapies that target these pathways.

Development of use-specific high-performance cyber-nanomaterial optical detectors by effective choice of machine learning algorithms

Due to their inherent variabilities, nanomaterials-based sensors are challenging to translate into real-world applications, where reliability and reproducibility are key. Machine learning can be a powerful approach for obtaining reliable inferences from data generated by such sensors. Here, we show that the best choice of ML algorithm in a cyber-nanomaterial detector is largely determined by the specific use-considerations, including accuracy, computational cost, speed, and resilience against drifts and long-term ageing effects. When sufficient data and computing resources are provided, the highest sensing accuracy can be achieved by the k-nearest neighbors (kNNs) and Bayesian inference algorithms, however, these algorithms can be computationally expensive for real-time applications. In contrast, artificial neural networks (ANNs) are computationally expensive to train (off-line), but they provide the fastest result under testing conditions (on-line) while remaining reasonably accurate. When access to data is limited, support vector machines (SVMs) can perform well even with small training sample sizes, while other algorithms show considerable reduction in accuracy if data is scarce, hence, setting a lower limit on the size of required training data. We also show by tracking and modeling the long-term drifts of the detector performance over a one year time-frame, it is possible to dramatically improve the predictive accuracy without any re-calibration. Our research shows for the first time that if the ML algorithm is chosen specific to the use-case, low-cost solution-processed cyber-nanomaterial detectors can be practically implemented under diverse operational requirements, despite their inherent variabilities.

Influenza Prediction: Analyzing Machine Learning Algorithms

Analyzing online or digital data for detecting epidemics is one of the hot areas of research and now becomes more relevant during the present outbreak of Covid-19. There are several different types of the influenza virus and moreover they keep evolving constantly in the same manner the COVID-19 virus has done. As a result, they pose a greater challenge when it comes to analyzing them, predicting when, where and at what degree of severity it will outbreak during the flu season across the world. There is need for greater surveillance to both seasonal and pandemic influenza to ensure the health and safety of the mankind. The objective of work is to apply machine learning algorithms for building predictive models that can predict where the occurrence, peak and severity of influenza in each season. For this work we have considered a freely available dataset of Ireland which is recorded for the duration of 2005 to 2016. Specifically, we have tested three ML Algorithms namely Linear Regression, Support Vector Regression and Random Forests. We found Random Forests is giving better predictive results. We also conducted experiment through weka tool and tested Zero R, Linear Regression, Lazy Kstar, Random Forest, REP Tree, Multilayer Perceptron models. We again found the Random Forest is performing better in comparison to all other models. We also evaluated other regression models including Ridge Regression, modified Ridge regression, Lasso Regression, K Neighbor Regression and evaluated the mean absolute errors. We found that modified Ridge regression is producing minimum error. The proposed work is inclined towards finding the suitability & appropriate ML algorithm for solving this problem on Flu.

Abstract A32: Identifying actionable pathway malfunction scores with ML algorithm for omics data

Abstract Background: Driver mutations are traditionally considered as actionable biomarkers for targeted drugs, but the resistance and relapse effects often occur even when these events are precisely discovered. At the same time, primary DNA mutations can be only the triggers for cell malignancy and further development of the tumor occurs due to following pathways imbalance, which may be reflected in gene expression. The goal is to detect preaffected pathways that are most close to the oncogenic affected state, so during the treatment strategy planning we could consider these pathways as the next potential targets after nonresponse or relapse. Methods: We took the data from TCGA Pan-Cancer Atlas on whole-exome sequencing and RNA expression for 33 cancers. Mutations were filtered based on their pathogenicity (1,2). The training set included data on mutations and corresponded RNA levels of 1821 cancer pathways-related genes (3). ML method-logistic regression, with 5-fold cross-validation with a test set, was realized on Python 3.7. Results: Using gene expression data, 9 most common actionable events were predicted: oncogenic mutations affecting Ras, Raf, Ras/Raf/MEK, PI3K, CDK protein families, amplifications of EGFR, ERBB2, CDK4 genes, with an accuracy of 80% - 93%. Results were the probabilities of events: range 10-30% occurrence is shown. Discussion: We considered the obtained molecular events probabilities as the scores of corresponding pathways’ malfunctions. For some molecular events, more than one-third of patients has >10% affected (unbalanced) pathway state. This approach after validation can be used in clinical research practice for patient cohorts risk stratification, or as additional reinforcement for drug companion tests.

Machine learning analysis of exome trios to contrast the genomic architecture of autism and schizophrenia

BackgroundMachine learning (ML) algorithms and methods offer great tools to analyze large complex genomic datasets. Our goal was to compare the genomic architecture of schizophrenia (SCZ) and autism spectrum disorder (ASD) using ML.MethodsIn this paper, we used regularized gradient boosted machines to analyze whole-exome sequencing (WES) data from individuals SCZ and ASD in order to identify important distinguishing genetic features. We further demonstrated a method of gene clustering to highlight which subsets of genes identified by the ML algorithm are mutated concurrently in affected individuals and are central to each disease (i.e., ASD vs. SCZ “hub” genes).ResultsIn summary, after correcting for population structure, we found that SCZ and ASD cases could be successfully separated based on genetic information, with 86–88% accuracy on the testing dataset. Through bioinformatic analysis, we explored if combinations of genes concurrently mutated in patients with the same condition (“hub” genes) belong to specific pathways. Several themes were found to be associated with ASD, including calcium ion transmembrane transport, immune system/inflammation, synapse organization, and retinoid metabolic process. Moreover, ion transmembrane transport, neurotransmitter transport, and microtubule/cytoskeleton processes were highlighted for SCZ.ConclusionsOur manuscript introduces a novel comparative approach for studying the genetic architecture of genetically related diseases with complex inheritance and highlights genetic similarities and differences between ASD and SCZ.

Machine Learning to Predict Stent Restenosis Based on Daily Demographic, Clinical, and Angiographic Characteristics

BackgroundMachine learning (ML) has arrived in medicine to deliver individually adapted medical care. This study sought to use ML to discriminate stent restenosis (SR) compared with existing predictive scores of SR. To develop an easily applicable model, we performed our predictions without any additional variables other than those obtained in daily practice. MethodsThe dataset, obtained from the Grupo de Análisis de la Cardiopatía Isquémica Aguda (GRACIA)-3 trial, consisted of 263 patients with demographic, clinical, and angiographic characteristics; 23 (9%) of them presented with SR at 12 months after stent implantation. A methodology to work with small imbalanced datasets, based in cross-validation and the precision/recall (PR) plots, was used, and state-of-the-art ML classifiers were trained. ResultsOur best performing model (0.46, area under the PR curve [AUC-PR]) was developed with an extremely randomized trees classifier, which showed better performance than chance alone (0.09 AUC-PR, corresponding to the 9% of patients presenting SR in our dataset) and 3 existing scores; Prevention of Restenosis With Tranilast and its Outcomes (PRESTO)-1 (0.31 AUC-PR), PRESTO-2 (0.27 AUC-PR), and Evaluation of Drug-Eluting Stents and Ischemic Events (EVENT) (0.18 AUC-PR). The most important variables ranked according to their contribution to the predictions were diabetes, ≥2 vessel-coronary disease, post-percutaneous coronary intervention thrombolysis in myocardial infarction (PCI TIMI)-flow, abnormal platelets, post-PCI thrombus, and abnormal cholesterol. To counteract the lack of external validation for our study, we deployed our ML algorithm in an open source calculator, in which the model would stratify patients of high and low risk as an example tool to determine generalizability of prediction models from small imbalanced sample size. ConclusionsApplied immediately after stent implantation, a ML model better differentiates those patients who will present with SR over current discriminators.

Optimization Method of Node Location Accuracy Based on Artificial Intelligence Technology

Wireless sensor networks(WSN) are one of the hot topics in current information technology research. With the diversification of wireless sensor network applications and the gradual maturity of node positioning technology for static networks, mobile node positioning technology. In order to find a method for optimizing the positioning accuracy of wireless network nodes based on artificial intelligence technology and improve the positioning accuracy, a large-scale simulation experiment is performed in this paper. After constructing a probabilistic model and conducting a large number of simulation experiments, a series of random variable samples are obtained. After statistical processing of these simulation samples, a solution to the problem is found accordingly. By comparing the average positioning error with the ML algorithm, it is found that the SA-L algorithm using artificial intelligence method has less dependence on the density of beacon nodes, improves the accuracy of network node positioning, and is more suitable for node positioning of WSN. Research shows that artificial intelligence technology can improve the accuracy of network node positioning and provide a lot of reference for the development of wireless networks.

Intelligent Vehicle Speed Controlling and Pothole Detection System

Speed is the single largest killer on India’s roads. The higher the speed, the greater the impact and the more the chances of grievous injury and death. This is why speed management is something which needs to be seriously considered. Here is an approach to build a system which reduces the number of accidents due to driver’s negligence of over speeding. We are proposing a Dynamic Speed Limiter with the help of machine learning algorithm which will help in reducing the accidents caused due to over speeding and rash driving of vehicles on road. The proposed system set the maximum speed of vehicle with the help of sign board available on the roads which define the safe driving speed of vehicle for that area or road. Also to solve the potholes we are using ML algorithm and mobile phone accelerometer to detect the potholes, accelerometer vibrations are set. After detecting the vibrations, the location (i.e. longitude and latitude) is marked on the GPS, commuters will get information about potholes.

TCAD-Machine Learning Framework for Device Variation and Operating Temperature Analysis With Experimental Demonstration

This work, for the first time, experimentally demonstrates a TCAD-Machine Learning (TCAD-ML) framework to assist the analysis of device-to-device variation and operating (ambient) temperature without the need of physical quantities extraction. The ML algorithm used in this work is the Principal Component Analysis (PCA) followed by third order polynomial regression. After calibrated to limited ‘expensive’ experimental data, ‘low cost’ TCAD simulation is used to generate a large amount of device data to train the ML model. The ML was then used to identify the root cause of device variation and operating temperature from any given experimental current-voltage (I-V) characteristics. We applied this framework to study the ultra-wide-bandgap gallium oxide (Ga2O3) Schottky barrier diode (SBD), an emerging device technology that holds great promise for temperature sensing, RF, and power applications in harsh environments. After calibration, over 150,000 electrothermal TCAD simulations are performed with random variation of physical parameters (anode effective work function, drift layer doping, and drift layer thickness) and operating temperature. An ML model is trained using these TCAD data and we found 1,000–10,000 TCAD data can train an accurate machine. We show that without physical quantities extraction, performing PCA is essential for the TCAD trained ML model to be applicable to analyze experimental characteristics. The physical parameters and temperatures predicted by the ML model show good agreement with experimental analysis. Our TCAD-ML framework shows great promise to accelerate the development of new device technologies with a significantly more efficient process of material and device experimentation.

A Note on the Interpretability of Machine Learning Algorithms

We are interested in the analysis of the concept of interpretability associated with a ML algorithm. We distinguish between the “How”, i.e., how a black box or a very complex algorithm works, and the “Why”, i.e. why an algorithm produces such a result. These questions appeal to many actors, users, professions, regulators among others. Using a formal standardized framework, we indicate the solutions that exist by specifying which elements of the supply chain are impacted when we provide answers to the previous questions. This presentation, by standardizing the notations, allows to compare the different approaches and to highlight the specificities of each of them: both their objective and their process. The study is not exhaustive and the subject is far from being closed.

Early Exposure of Lung Cancer by Combining ANN and SVM Algorithms

Lung cancer is a lethal type of cancers as its rate of spreading is very high compared to the other cancers. Patient who have been affected from Small Cell Lung Cancer (SCLC) has fast outspread rate. Even at initial stage, around 67-75% of cancer victims with SCLC will have fast outspreads and serious damages to the nearby physical parts. Moreover, World Health Organization (WHO) has predicted the count of lung cancer deaths will reach 9.6 million in 2020. Identifying such a lethal type of cancer early can be lifesaving one. Because, cancer cells in lungs are capable of traveling to other body parts even before the doctor detects them in lungs. In this research work, we have designed a combined approach to prognosticate lung cancer and its type using Artificial Neural Networks (ANN) and Support Vector Machine (SVM). To train both the ML algorithm, an open access patient health dataset published by cancer imaging archives is used. The dataset has the information like pretreatment CT scans, 3D image details of tumor and clinical outcomes. The results produced by ANN and SVM algorithm are compared to predict the type of the lung cancer accurately. The result holds good for a real time implementation.

Does Skillful Use of Hard Information by Machines Outperform a Combination of Hard and Soft Information of Loan Officers in Lending Decisions?

We ask whether machine learning algorithms improve the efficiency of screening of the loan officers, and thereby help expand access to formal credit. We obtain loan application level data from an Indian bank. To overcome the selective labels problem, we exploit the incentive driven within officer difference in leniency within a calendar month. We find that the ML algorithm is able to lend 26.6% more at loan officers' delinquency rate or achieve 21.3% lower delinquency rate at loan officers' approval rate. Higher efficiency is achieved without compromising on equity.

30Machine learning to predict the long-term risk of myocardial infarction and cardiac death based on clinical risk, coronary calcium and epicardial adipose tissue: a prospective study

Abstract Background/Introduction Machine learning (ML) allows objective integration of clinical and imaging data for the prediction of events. ML prediction of cardiovascular events in asymptomatic subjects over long-term follow-up, utilizing quantitative CT measures of coronary artery calcium (CAC) and epicardial adipose tissue (EAT) have not yet been evaluated. Purpose To analyze the ability of machine learning to integrate clinical parameters with coronary calcium and EAT quantification in order to improve prediction of myocardial infarction (MI) and cardiac death in asymptomatic subjects. Methods We assessed 2071 consecutive subjects [1230 (59%) male, age: 56.049.03] from the EISNER (Early Identification of Subclinical Atherosclerosis by Noninvasive Imaging Research) trial with long-term follow-up after non-enhanced cardiac CT. CAC (Agatston) score, age-and-gender-adjusted CAC percentile, and aortic calcium scores were obtained. EAT volume and density were quantified using a fully automated deep learning method. Extreme gradient boosting, a ML algorithm, was trained using demographic variables, plasma lipid panel measurements, risk factors as well as CAC, aortic calcium and EAT measures from CAC CT scans. ML was validated using 10-fold cross validation; event prediction was evaluated using area-under-receiver operating characteristic curve (AUC) analysis and Cox proportional hazards regression. Optimal ML cut-point for risk of MI and cardiac death was determined by highest Youden's index (sensitivity + specificity – 1). Results At 152 years' follow-up, 76 events of MI and/or cardiac death had occurred. ML obtained a significantly higher AUC than the ASCVD risk and CAC score in predicting events (ML: 0.81; ASCVD: 0.76, p<0.05; CAC: 0.75, p<0.01, Figure A). ML performance was mostly driven by age, ASCVD risk and calcium as shown by the variable importance (Figure B); however, all variables with non-zero gain contributed to the ML performance. ML achieved a sensitivity and specificity of 77.6% and 73.5%, respectively. For an equal specificity, ASCVD and CAC scores obtained a sensitivity of 61.8% and 67.1%, respectively. High ML risk was associated with a high risk of suffering an event by Cox regression (HR: 9.25 [95% CI: 5.39–15.87], p<0.001; survival curves in Figure C). The relationships persisted when adjusted for age, gender, CAC, CAC percentile, aortic calcium score, and ASCVD risk score; with a hazard ratio of 3.42 for high ML risk (HR: 3.42 [95% CI: 1.54–7.57], p=0.002). Conclusion(s) Machine learning used to integrate clinical and quantitative imaging-based variables significantly improves prediction of MI and cardiac death in asymptomatic subjects undergoing CAC assessment, compared to standard risk assessment methods. Acknowledgement/Funding NHLBI 1R01HL13361, Bundesministerium für Bildung und Forschung (01EX1012B), Dr. Miriam and Sheldon G. Adelson Medical Research Foundation

Artificial Neural Network-Based Variable Importance Analysis of Prognostic Factors Related to Radiation Pneumonitis in Patients with Lung Cancer: Preliminary Study

Radiation pneumonitis (RP) is a major radiation-induced lung injury in patients with lung cancer. When an extremely high risk of severe RP is predicted, it is a critical issue that severely affects radiation dose and treatment planning. It is essential to assess which prognostic factor affects the occurrence of RP before initiating radiation treatment. In this study, we aimed to identify the variable importance (VI) of prognostic factors related to RP by using an artificial neural network (ANN) containing complex association between several prognostic factors. We reviewed 110 cases in patients with lung cancer who received radiation therapy (RT) from August 2000 to December 2018. The multi-layer perceptron algorithm, which is a back-propagation ANN algorithm, was implemented by SPSS Modeler (verl3.1, IBM SPSS Inc., Chicago, IL). The fifteen input variables were set, and the target variable was the occurrence of RP. The VI, which indicates the effect of each prognostic factor on the occurrence of RP, was analyzed by the variance-based method. To evaluate the VI of the ANN, we qualitatively compared the VI of the ANN with the odds ratios (OR) obtained from previously published literature. Patients who had an RP grade >2 comprised 13.6%. The accuracy of the ANN model was 76.92% and the and the area under the curve (AUC) was 0.774. From the results of the VI from the ANN, mean lung dose (MLD) had the highest VI among the prognostic factors at 15.96%. The OR of interstitial lung disease (ILD) was the highest at 25.70. While the VI of the ANN includes complex associations of the prognostic factors, ORs include independent associations between exposure and outcome. This preliminary study is meaningful as it proposes a method to quantify the VI of prognostic factors related to patient outcome using even simple ML algorithm. By further acquiring the more refined big data, optimizing the ML model, and performing clinical validation, the accurate VI can be adopted for personalized RT planning.

A Reliable Workload Management Based on Predictive Analysis and Characterization of Workload Resources in HPC

Execution of Big Data workloads upon High Performance Computing (HPC) infrastructures has become an attractive way to improve their performances. In resource management, a large volume of multi-structured log data of Cloud Data Center (CDC) is generated regarding job arrival patterns, CPU memory consumption, task duration and many others. The system also provides the mechanism for CACO Cauchy matrix method for automatic data recovery from disk failure; it can also remove the lengthy process like replica management. In this paper we proposed a system for dynamic load rebalancing and resource allocation technique using machine learning algorithm. Q-Learning based on ML algorithm has used for validating the system. Experimental analysis illustrates that how proposed system eliminates the present approaches drawbacks. Our key is to identify the various workload patterns generation in heterogeneous storage environments using machine learning algorithm.