RESUMEN

Stroke is a severe illness, that requires early stroke detection and intervention, as this would help prevent the worsening of the condition. The research is done to solve stroke prediction problem, which may be divided into a number of sub-problems such as an individual's predisposition to develop stroke. To attain this objective, a multiturn dataset consisting of various health features, such as age, gender, hypertension, and glucose levels, takes a central role. A multiple approach was put forward concentrating on integrating the machine learning techniques, such as Logistic Regression, Naive Bayes, K-Nearest Neighbors, and Support Vector Machine (SV), together to develop an ensemble machine called Neuro-Health Guardian. The hypothesis "Neuro-Health Guardian Model" integrates these algorithms into one, purported to make stroke prediction more accurate. The topic dives into each instance of preparation of data for analysis, data visualization techniques, selection of the right model, training, testing, ensembling, evaluation, and prediction. The models are validated with error rate accounted from their accuracy, precision, recall, F1 score, and finally confusion matrices for a look. The study's result is showing that the ensemble model that combines the multiple algorithms has the edge over them and this is evidently by the fact that it can predict stroke rises. Additionally, accuracy, precision, recall, and F1 scores are measured in all models and the comparison is done to provide a clear comparison of the models' performance. In short, the article presented the formation of the ongoing stroke prediction that revealed the ensemble model as a good anticipation. Precise stroke predisposition forecasting can assist in early intervention thereby preventing stroke-related deaths, and limiting disability burden by stroke. The conclusions that have come out of this study offer a great action item for the development of predictive models related to stroke prevention and treatment.

Asunto(s)

RESUMEN

The high sensitivity and picosecond time resolution of single-photon avalanche diodes (SPADs) can improve the operational range and imaging accuracy of underwater detection systems. When an underwater SPAD imaging system is used to detect targets, backward-scattering caused by particles in water often results in the poor quality of the reconstructed underwater image. Although methods such as simple pixel accumulation have been proven to be effective for time-photon histogram reconstruction, they perform unsatisfactorily in a highly scattering environment. Therefore, new reconstruction methods are necessary for underwater SPAD detection to obtain high-resolution images. In this paper, we propose an algorithm that reconstructs high-resolution depth profiles of underwater targets from a time-photon histogram by employing the K-nearest neighbor (KNN) to classify multiple targets and the background. The results contribute to the performance of pixel accumulation and depth estimation algorithms such as pixel cross-correlation and ManiPoP. We use public experimental data sets and underwater simulation data to verify the effectiveness of the proposed algorithm. The results of our algorithm show that the root mean square errors (RMSEs) of land targets and simulated underwater targets are reduced by 57.12% and 23.45%, respectively, achieving high-resolution single-photon depth profile reconstruction.

RESUMEN

Background: Tandem duplication (TD) is a common and important type of structural variation in the human genome. TDs have been shown to play an essential role in many diseases, including cancer. However, it is difficult to accurately detect TDs due to the uneven distribution of reads and the inherent complexity of next-generation sequencing (NGS) data. Methods: This article proposes a method called DTDHM (detection of tandem duplications based on hybrid methods), which utilizes NGS data to detect TDs in a single sample. DTDHM builds a pipeline that integrates read depth (RD), split read (SR), and paired-end mapping (PEM) signals. To solve the problem of uneven distribution of normal and abnormal samples, DTDHM uses the K-nearest neighbor (KNN) algorithm for multi-feature classification prediction. Then, the qualified split reads and discordant reads are extracted and analyzed to achieve accurate localization of variation sites. This article compares DTDHM with three other methods on 450 simulated datasets and five real datasets. Results: In 450 simulated data samples, DTDHM consistently maintained the highest F1-score. The average F1-score of DTDHM, SVIM, TARDIS, and TIDDIT were 80.0%, 56.2%, 43.4%, and 67.1%, respectively. The F1-score of DTDHM had a small variation range and its detection effect was the most stable and 1.2 times that of the suboptimal method. Most of the boundary biases of DTDHM fluctuated around 20 bp, and its boundary deviation detection ability was better than TARDIS and TIDDIT. In real data experiments, five real sequencing samples (NA19238, NA19239, NA19240, HG00266, and NA12891) were used to test DTDHM. The results showed that DTDHM had the highest overlap density score (ODS) and F1-score of the four methods. Conclusions: Compared with the other three methods, DTDHM achieved excellent results in terms of sensitivity, precision, F1-score, and boundary bias. These results indicate that DTDHM can be used as a reliable tool for detecting TDs from NGS data, especially in the case of low coverage depth and tumor purity samples.

Asunto(s)

RESUMEN

The preliminary classification of biological class data is of great importance for bioinformatics. One can quickly classify object data by comparing their existing features with known traits. k-nearest neighbor algorithm is easy to apply in this field, but its drawbacks make it less meaningful to improve the efficiency of the algorithm by simply changing the distance model, so this study uses a local mean-based k-nearest neighbor classifier and compares the accuracy of the predicted classification of six different distance models used. The prediction accuracies in the experimental results were all greater than 70%, and the highest accuracy was achieved in different data sets for all distance models with K=2; the prediction accuracy of Minkowski distance with different parameters had the highest volatility in the test.and the experimental results can be used as a reference for related practitioners.

Asunto(s)

RESUMEN

In order to stop deterioration and give patients with Alzheimer's disease (AD) early therapy, it is crucial to correctly diagnose AD and its early stage, mild cognitive impairment (MCI). A framework for diagnosing AD is presented in this paper, which includes magnetic resonance imaging (MRI) image preprocessing, feature extraction, and the Fuzzy k-nearest neighbor algorithm (FKNN) model. In particular, the framework's novelty lies in the use of an improved Harris Hawks Optimization (HHO) algorithm named SSFSHHO, which integrates the Sobol sequence and Stochastic Fractal Search (SFS) mechanisms for optimizing the parameters of FKNN. The HHO method improves the quality of the initial population overall by incorporating the Sobol sequence, and the SFS mechanism increases the algorithm's capacity to get out of the local optimum solution. Comparisons with other classical meta-heuristic algorithms, state-of-the-art HHO variants in low and high dimensions, and enhanced meta-heuristic algorithms on 30 typical IEEE CEC2014 benchmark test problems show that the overall performance of SSFSHHO is significantly better than other comparative algorithms. Moreover, the created framework based on the SSFSHHO-FKNN model is employed to classify AD and MCI using MRI scans from the ADNI dataset, achieving high classification performance for 6 representative cases. Experimental findings indicate that the proposed algorithm performs better than a number of high-performance optimization algorithms and classical machine learning algorithms, thus offering a promising approach for AD classification. Additionally, the proposed strategy can successfully identify relevant features and enhance classification performance for AD diagnosis.

Asunto(s)

RESUMEN

The K-nearest neighbor (KNN) algorithm is one of the most extensively used classification algorithms, while its high time complexity limits its performance in the era of big data. The quantum K-nearest neighbor (QKNN) algorithm can handle the above problem with satisfactory efficiency; however, its accuracy is sacrificed when directly applying the traditional similarity measure based on Euclidean distance. Inspired by the Polar coordinate system and the quantum property, this work proposes a new similarity measure to replace the Euclidean distance, which is defined as Polar distance. Polar distance considers both angular and module length information, introducing a weight parameter adjusted to the specific application data. To validate the efficiency of Polar distance, we conducted various experiments using several typical datasets. For the conventional KNN algorithm, the accuracy performance is comparable when using Polar distance for similarity measurement, while for the QKNN algorithm, it significantly outperforms the Euclidean distance in terms of classification accuracy. Furthermore, the Polar distance shows scalability and robustness superior to the Euclidean distance, providing an opportunity for the large-scale application of QKNN in practice.

RESUMEN

In this study, the predictive power of three different machine learning (ML)-based approaches, namely, multi-gene genetic programming (MGGP), M5 model trees (M5Tree), and K-nearest neighbor algorithm (KNN), for long-term monthly reference evapotranspiration (ET0) prediction were investigated. The input data consist of monthly solar radiation (Rs), maximum air temperature (Tmax), and wind speed (Ws) derived from 163 meteorological stations in Turkey. Different input combinations were created and analyzed. The model's performance was evaluated using criteria such as Nash-Sutcliffe efficiency, Kling-Gupta efficiency, relative root mean squared error, mean absolute percentage error, and determination coefficient. Moreover, Taylor, radar, and boxplot diagrams were created. It was determined that the MGGP model outperformed both the M5Tree and the KNN models. The equation obtained from the MGGP model, for the best-performed combination of Rs-Tmax-Ws, was presented. The best weather conditions were obtained as 0.029 to 31.814 MJ/m2, - 5.8 to 45.7 °C, and 0.140 to 5.086 m/s for Rs, Tmax, and Ws, respectively. It was also found that the Rs was the most potent input variable for ET0 estimation while Ws was the weakest.

Asunto(s)

RESUMEN

Cryptographic algorithm identification, which refers to analyzing and identifying the encryption algorithm used in cryptographic system, is of great significance to cryptanalysis. In order to improve the accuracy of identification work, this article proposes a new ensemble learning-based model named hybrid k-nearest neighbor and random forest (HKNNRF), and constructs a block cipher algorithm identification scheme. In the ciphertext-only scenario, we use NIST randomness test methods to extract ciphertext features, and carry out binary-classification and five-classification experiments on the block cipher algorithms using proposed scheme. Experiments show that when the ciphertext size and other experimental conditions are the same, compared with the baselines, the HKNNRF model has higher classification accuracy. Specifically, the average binary-classification identification accuracy of HKNNRF is 69.5%, which is 13%, 12.5%, and 10% higher than the single-layer support vector machine (SVM), k-nearest neighbor (KNN), and random forest (RF) respectively. The five-classification identification accuracy can reach 34%, which is higher than the 21% accuracy of KNN, the 22% accuracy of RF and the 23% accuracy of SVM respectively under the same experimental conditions.

RESUMEN

Recently, indoor localization has become an active area of research. Although there are various approaches to indoor localization, methods that utilize artificially generated magnetic fields from a target device are considered to be the best in terms of localization accuracy under non-line-of-sight conditions. In magnetic field-based localization, the target position must be calculated based on the magnetic field information detected by multiple sensors. The calculation process is equivalent to solving a nonlinear inverse problem. Recently, a machine-learning approach has been proposed to solve the inverse problem. Reportedly, adopting the k-nearest neighbor algorithm (k-NN) enabled the machine-learning approach to achieve fairly good performance in terms of both localization accuracy and computational speed. Moreover, it has been suggested that the localization accuracy can be further improved by adopting artificial neural networks (ANNs) instead of k-NN. However, the effectiveness of ANNs has not yet been demonstrated. In this study, we thoroughly investigated the effectiveness of ANNs for solving the inverse problem of magnetic field-based localization in comparison with k-NN. We demonstrate that despite taking longer to train, ANNs are superior to k-NN in terms of localization accuracy. The k-NN is still valid for predicting fairly accurate target positions within limited training times.

Asunto(s)

RESUMEN

The security of water distribution systems has become the subject of an increasing volume of research over the last decade. Data analysis and machine learning are linked to hydraulic and quality modeling for improving the capacity of water utilities to save lives when faced with the contamination of water networks. This research applies k-nearest neighbor and random forest algorithms to estimate the location of contamination sources at near-real time. Epanet and Epanet-MSX software are used to simulate intrusions of pesticide into water distribution system and the interaction with compounds already present in water bulk. Different pesticide concentrations are considered in the simulations, and chlorine monitoring occurs through placed quality sensors. The results show that random forest can localize [Formula: see text] of contamination scenarios, while the KNN algorithm found [Formula: see text]. Finally, an assessment of contamination spread is made for a better understanding of the impacts of non-localized contamination.

Asunto(s)

RESUMEN

American foulbrood is a dangerous disease of bee broods found worldwide, caused by the Paenibacillus larvae larvae L. bacterium. In an experiment, the possibility of detecting colonies of this bacterium on MYPGP substrates (which contains yeast extract, Mueller-Hinton broth, glucose, K2HPO4, sodium pyruvate, and agar) was tested using a prototype of a multi-sensor recorder of the MCA-8 sensor signal with a matrix of six semiconductors: TGS 823, TGS 826, TGS 832, TGS 2600, TGS 2602, and TGS 2603 from Figaro. Two twin prototypes of the MCA-8 measurement device, M1 and M2, were used in the study. Each prototype was attached to two laboratory test chambers: a wooden one and a polystyrene one. For the experiment, the strain used was P. l. larvae ATCC 9545, ERIC I. On MYPGP medium, often used for laboratory diagnosis of American foulbrood, this bacterium produces small, transparent, smooth, and shiny colonies. Gas samples from over culture media of one- and two-day-old foulbrood P. l. larvae (with no colonies visible to the naked eye) and from over culture media older than 2 days (with visible bacterial colonies) were examined. In addition, the air from empty chambers was tested. The measurement time was 20 min, including a 10-min testing exposure phase and a 10-min sensor regeneration phase. The results were analyzed in two variants: without baseline correction and with baseline correction. We tested 14 classifiers and found that a prototype of a multi-sensor recorder of the MCA-8 sensor signal was capable of detecting colonies of P. l. larvae on MYPGP substrate with a 97% efficiency and could distinguish between MYPGP substrates with 1-2 days of culture, and substrates with older cultures. The efficacy of copies of the prototypes M1 and M2 was shown to differ slightly. The weighted method with Canberra metrics (Canberra.811) and kNN with Canberra and Manhattan metrics (Canberra. 1nn and manhattan.1nn) proved to be the most effective classifiers.

Asunto(s)

RESUMEN

OBJECTIVE: Air pollution has potential risk on asthma patients, further prolongs the length of stay. However, it is unclear that the impact of air pollution on excessive length of stay (ELoS) of heterogeneous asthma patients. In this study, we proposed a K-Nearest Neighbor (KNN) embedded approach incorporating with patient status to analyze the impact of short-term air pollution on the ELoS of asthma patients. METHODS: The KNN embedded approach includes two stages. Firstly, the KNN algorithm was employed to search for the most similar patient community and approximate kernel proxy of each index patient by Euclidean distance. Then, we built the differential fixed-effect linear model to estimate the risk of air pollution to the ELoS. RESULTS: We analyzed 6563 asthma patients' medical insurance records in a large city of China from January to December in 2014. It was found that when the duration of exposure to air pollution (i.e., PM2.5, PM10, SO2, NO2, and CO) reaches around 4-5 days, the risk of increasing the ELoS becomes the largest. But only O3 shows the opposite effect. What's more, CO is the dominant risk to increase the ELoS. With a 1 mg/m3 increment of CO average concentration in 5 days, the ELoS will go up by 0.8157 day (95%CI:0.72,0.9114). Based on the kernel proxy in the top 1% similar patient community, the additional financial burden posed on each patient increases by RMB 488.6002 (95%CI:430.1962,547.0043) due to the ELoS. CONCLUSIONS: The KNN embedded approach is an innovative method that takes into account the heterogeneous patient status, and effectively estimates the impact of air pollution on the ELoS. It is concluded that air pollution poses adverse effects and additional financial burdens on asthma patients. Heterogeneous patients should adopt different strategies in health management to reduce the risk of increasing the ELoS due to air pollution, and improve the efficiency of medical resource utilization. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s40201-020-00584-8.

RESUMEN

BACKGROUND: Heart rate variability (HRV) and electrocardiogram (ECG)-derived respiration (EDR) have been used to detect sleep apnea (SA) for decades. The present study proposes an SA-detection algorithm using a machine-learning framework and bag-of-features (BoF) derived from an ECG spectrogram. METHODS: This study was verified using overnight ECG recordings from 83 subjects with an average apnea-hypopnea index (AHI) 29.63 (/h) derived from the Physionet Apnea-ECG and National Cheng Kung University Hospital Sleep Center database. The study used signal preprocessing to filter noise and artifacts, ECG time-frequency transformation using continuous wavelet transform (CWT), BoF feature generation, machine-learning classification using support vector machine (SVM), ensemble learning (EL), k-nearest neighbor (KNN) classification, and cross-validation. The time length of the spectrogram was set as 10 and 60 s to examine the required minimum spectrogram window time length to achieve satisfactory accuracy. Specific frequency bands of 0.1-50, 8-50, 0.8-10, and 0-0.8 Hz were also extracted to generate the BoF to determine the band frequency best suited for SA detection. RESULTS: The five-fold cross-validation accuracy using the BoF derived from the ECG spectrogram with 10 and 60 s time windows were 90.5% and 91.4% for the 0.1-50 Hz and 8-50 Hz frequency bands, respectively. CONCLUSION: An SA-detection algorithm utilizing BoF and a machine-learning framework was successfully developed in this study with satisfactory classification accuracy and high temporal resolution.

RESUMEN

One of the key elements in assessing traffic safety on the roads is the detection of asphalt conditions. In this paper, we propose an optical sensor based on GeSi nanocrystals embedded in SiO2 matrix that discriminates between different slippery road conditions (wet and icy asphalt and asphalt covered with dirty ice) in respect to dry asphalt. The sensor is fabricated by magnetron sputtering deposition followed by rapid thermal annealing. The photodetector has spectral sensitivity in the 360-1350 nm range and the signal-noise ratio is 102-103. The working principle of sensor setup for detection of road conditions is based on the photoresponse (photocurrent) of the sensor under illumination with the light reflected from the asphalt having different reflection coefficients for dry, wet, icy and dirty ice coatings. For this, the asphalt is illuminated sequentially with 980 and 1064 nm laser diodes. A database of these photocurrents is obtained for the different road conditions. We show that the use of both k-nearest neighbor and artificial neural networks classification algorithms enables a more accurate recognition of the class corresponding to a specific road state than in the case of using only one algorithm. This is achieved by comparing the new output sensor data with previously classified data for each algorithm and then by performing an intersection of the algorithms' results.

RESUMEN

With the rapid improvement of the perception and computing capacity of mobile devices such as smart phones, human activity recognition using mobile devices as the carrier has been a new research hot-spot. The inertial information collected by the acceleration sensor in the smart mobile device is used for human activity recognition. Compared with the common computer vision recognition, it has the following advantages: convenience, low cost, and better reflection of the essence of human motion. Based on the WISDM data set collected by smart phones, the inertial navigation information and the deep learning algorithm-convolutional neural network (CNN) were adopted to build a human activity recognition model in this paper. The K nearest neighbor algorithm (KNN) and the random forest algorithm were compared with the CNN network in the recognition accuracy to evaluate the performance of the CNN network. The classification accuracy of CNN model reached 92.73%, which was much higher than KNN and random forest. Experimental results show that the CNN algorithm model can achieve more accurate human activity recognition and has broad application prospects in predicting and promoting human health.

Asunto(s)

RESUMEN

Varroosis is a dangerous and difficult to diagnose disease decimating bee colonies. The studies conducted sought answers on whether the electronic nose could become an effective tool for the efficient detection of this disease by examining sealed brood samples. The prototype of a multi-sensor recorder of gaseous sensor signals with a matrix of six semiconductor gas sensors TGS 823, TGS 826, TGS 832, TGS 2600, TGS 2602, and TGS 2603 from FIGARO was tested in this area. There were 42 objects belonging to 3 classes tested: 1st class-empty chamber (13 objects), 2nd class-fragments of combs containing brood sick with varroosis (19 objects), and 3rd class-fragments of combs containing healthy sealed brood (10 objects). The examination of a single object lasted 20 min, consisting of the exposure phase (10 min) and the sensor regeneration phase (10 min). The k-th nearest neighbors algorithm (kNN)-with default settings in RSES tool-was successfully used as the basic classifier. The basis of the analysis was the sensor reading value in 270 s with baseline correction. The multi-sensor MCA-8 gas sensor signal recorder has proved to be an effective tool in distinguishing between brood suffering from varroosis and healthy brood. The five-time cross-validation 2 test (5 × CV2 test) showed a global accuracy of 0.832 and a balanced accuracy of 0.834. Positive rate of the sick brood class was 0.92. In order to check the overall effectiveness of baseline correction in the examined context, we have carried out additional series of experiments-in multiple Monte Carlo Cross Validation model-using a set of classifiers with different metrics. We have tested a few variants of the kNN method, the Naïve Bayes classifier, and the weighted voting classifier. We have verified with statistical tests the thesis that the baseline correction significantly improves the level of classification. We also confirmed that it is enough to use the TGS2603 sensor in the examined context.

Asunto(s)

RESUMEN

Changes in potential evapotranspiration will affect the surface ecology and environment of the land. Accurate and quick estimation of potential evapotranspiration will help to analyze environmental change. In this study, in combination with the canonical correlation analysis (CCA) and k-nearest neighbor algorithm (k-NN), a new method for calculating potential evapotranspiration (CCA-k-NN) based on self-optimizing nearest neighbor algorithm was proposed, in which less meteorological data were used for estimation. By analyzing the basic principles of CCA and k-NN and according to the requirement of estimating ET0, the CCA-k-NN method was constructed, and its basic principles and key steps were described. In this method, CCA algorithm was used to find the most relevant meteorological data for potential evapotranspiration, and the dimensionality of meteorological data for subsequent estimation of ET0 was reduced. Then, k-NN algorithm was used to estimate ET0. The Northwest of China was chosen as the research area to evaluate the applicability of this method. The 148 data stations in the region were divided into training datasets, testing datasets, and validation datasets. ET0 was estimated on three datasets using the proposed method, and the estimation accuracy of the CCA-k-NN method was evaluated with FAO-56 Penman-Monteith as a reference. The results show that the CCA-k-NN method maintains a high correlation with FAO-56 Penman-Monteith (correlation coefficient is greater than 0.9) and has a good estimation accuracy. RMSE and MAE are both less than 1 mm day-1, and the overall performance of NSCE is greater than 0.5, all of which reach the level of "applicable" and above. At the same time, the CCA-k-NN method has low time complexity O(n). Comparison of the results of the CCA-k-NN method with those of other empirical models showed that the CCA-k-NN method is more accurate and can be employed successfully in estimating ET0.

Asunto(s)

RESUMEN

With the rapid improvement of the perception and computing capacity of mobile devices such as smart phones, human activity recognition using mobile devices as the carrier has been a new research hot-spot. The inertial information collected by the acceleration sensor in the smart mobile device is used for human activity recognition. Compared with the common computer vision recognition, it has the following advantages: convenience, low cost, and better reflection of the essence of human motion. Based on the WISDM data set collected by smart phones, the inertial navigation information and the deep learning algorithm-convolutional neural network (CNN) were adopted to build a human activity recognition model in this paper. The K nearest neighbor algorithm (KNN) and the random forest algorithm were compared with the CNN network in the recognition accuracy to evaluate the performance of the CNN network. The classification accuracy of CNN model reached 92.73%, which was much higher than KNN and random forest. Experimental results show that the CNN algorithm model can achieve more accurate human activity recognition and has broad application prospects in predicting and promoting human health.

Asunto(s)

RESUMEN

Formylation is a type of post-translational modification that can occur on lysine sites, which plays an irreplaceable role in organism. To better understand the mechanism, it is necessary to identify formylation sites in proteins accurately. Computational method is popular because of its more convenience and higher speed than traditional experimental methods. However, no computational method has been proposed for prediction of lysine formylation. In this study, we developed a predictor named LFPred to identify lysine formylation sites using sequence features (including amino acid composition (AAC), binary profile features (BPF), and amino acid index (AAI)) combined K-nearest neighbor algorithm as classifier. We chose discrete window instead of continuous window according to information entropy. Besides, we took measure to select more reliable negative samples and address the severe imbalance between positive samples and negative samples. Finally, the performance of LFPred is measured with a specificity of 79.9% and a sensibility of 81.4% using jackknife test method, which indicated that our method can be a useful tool for prediction of lysine formylation sites.

Asunto(s)

RESUMEN

An electronic nose (e-nose) was used to characterize sesame oils processed by three different methods (hot-pressed, cold-pressed, and refined), as well as blends of the sesame oils and soybean oil. Seven classification and prediction methods, namely PCA, LDA, PLS, KNN, SVM, LASSO and RF, were used to analyze the e-nose data. The classification accuracy and MAUC were employed to evaluate the performance of these methods. The results indicated that sesame oils processed with different methods resulted in different sensor responses, with cold-pressed sesame oil producing the strongest sensor signals, followed by the hot-pressed sesame oil. The blends of pressed sesame oils with refined sesame oil were more difficult to be distinguished than the blends of pressed sesame oils and refined soybean oil. LDA, KNN, and SVM outperformed the other classification methods in distinguishing sesame oil blends. KNN, LASSO, PLS, and SVM (with linear kernel), and RF models could adequately predict the adulteration level (% of added soybean oil) in the sesame oil blends. Among the prediction models, KNN with k = 1 and 2 yielded the best prediction results.

Asunto(s)