Sequential Forward Feature Selection Research Articles

Enhancing phenotyping efficiency in faba bean breeding: integrating UAV imaging and machine learning

Unmanned aerial vehicles (UAVs) equipped with high-resolution imaging sensors have shown great potential for plant phenotyping in agricultural research. This study aimed to explore the potential of UAV-derived red–green–blue (RGB) and multispectral imaging data for estimating classical phenotyping measures such as plant height and predicting yield and chlorophyll content (indicated by SPAD values) in a field trial of 38 faba bean (Vicia faba L.) cultivars grown at four replicates in south-eastern Norway. To predict yield and SPAD values, Support Vector Regression (SVR) and Random Forest (RF) models were utilized. Two feature selection methods, namely the Pearson correlation coefficient (PCC) and sequential forward feature selection (SFS), were applied to identify the most relevant features for prediction. The models incorporated various combinations of multispectral bands, indices, and UAV-based plant height values at four different faba bean development stages. The correlation between manual and UAV-based plant height measurements revealed a strong agreement with a correlation coefficient (R2) of 0.97. The best prediction of SPAD value was achieved at BBCH 50 (flower bud present) with an R2 of 0.38 and RMSE of 1.14. For yield prediction, BBCH 60 (first flower open) was identified as the optimal stage, using spectral indices yielding an R2 of 0.83 and RMSE of 0.53 tons/ha. This development stage presents an opportunity to implement targeted management practices to enhance yield. The integration of UAVs equipped with RGB and multispectral cameras, along with machine learning algorithms, proved to be an accurate approach for estimating agronomically important traits in faba bean. This methodology offers a practical solution for rapid and efficient high-throughput phenotyping in faba bean breeding programs.

A hybrid machine learning-based model for predicting flight delay through aviation big data

The prediction of flight delays is one of the important and challenging issues in the field of scheduling and planning flights by airports and airlines. Therefore, in recent years, we have witnessed various methods to solve this problem using machine learning techniques. In this article, a new method is proposed to address these issues. In the proposed method, a group of potential indicators related to flight delay is introduced, and a combination of ANOVA and the Forward Sequential Feature Selection (FSFS) algorithm is used to determine the most influential indicators on flight delays. To overcome the challenges related to large flight data volumes, a clustering strategy based on the DBSCAN algorithm is employed. In this approach, samples are clustered into similar groups, and a separate learning model is used to predict flight delays for each group. This strategy allows the problem to be decomposed into smaller sub-problems, leading to improved prediction system performance in terms of accuracy (by 2.49%) and processing speed (by 39.17%). The learning model used in each cluster is a novel structure based on a random forest, where each tree component is optimized and weighted using the Coyote Optimization Algorithm (COA). Optimizing the structure of each tree component and assigning weighted values to them results in a minimum 5.3% increase in accuracy compared to the conventional random forest model. The performance of the proposed method in predicting flight delays is tested and compared with previous research. The findings demonstrate that the proposed approach achieves an average accuracy of 97.2% which indicates a 4.7% improvement compared to previous efforts.

A data driven approach to assess the petrophysical parametric sensitivity for lithology identification based on ensemble learning

Lithology identification is essential for geological analysis and reservoir characterization of hydrocarbon bearing formation. Conventional approaches of lithology identification are not only inconvenient but also time-consuming and expensive. Machine learning (ML) algorithms have been introduced as an effective, reliable, and accurate method of lithology identification using well-log data due to their ability to recognize the nonlinear relationship between log responses and lithology. In this study, we employed the XGBoost classifier model to assess the parametric sensitivity of the logging features in identifying lithology. Sequential Forward Feature Selection (SFFS) approach was applied to determine the most significant set of logging features that yields maximum classification accuracy. Nine logging parameters were used as the input of the XGBoost model to classify five lithology classes. Hyper-parameters were tuned within appropriate tuning search range to optimize the model performance along with Grid Search and 5-fold cross validation. The training and testing accuracy obtained from the model are 96% and 90.97%, respectively. The optimized XGBoost model showed the lowest precision, recall, and F1 score for the clay class. The shale class, on the other side, had the highest precision, recall, and F1 score of 95%, 94%, and 94%, respectively. The best set of logging features for identifying formation lithology is determined to be [PEF, SP, LLD, CALI, MSFL, NPHI, LLS], with a classification accuracy of 92.10% for the testing dataset. The least significant logging parameters for identifying lithology were found to be DT and GR, as their addition with the best input feature set, resulted in a negative deviation in classification accuracy. Finally, based on the classification accuracy, we ranked the top three sets of logging features, those contributed significantly.

Machine learning based biomarker discovery for chronic kidney disease-mineral and bone disorder (CKD-MBD).

Chronic kidney disease-mineral and bone disorder (CKD-MBD) is characterized by bone abnormalities, vascular calcification, and some other complications. Although there are diagnostic criteria for CKD-MBD, in situations when conducting target feature examining are unavailable, there is a need to investigate and discover alternative biochemical criteria that are easy to obtain. Moreover, studying the correlations between the newly discovered biomarkers and the existing ones may provide insights into the underlying molecular mechanisms of CKD-MBD. We collected a cohort of 116 individuals, consisting of three subtypes of CKD-MBD: calcium abnormality, phosphorus abnormality, and PTH abnormality. To identify the best biomarker panel for discrimination, we conducted six machine learning prediction methods and employed a sequential forward feature selection approach for each subtype. Additionally, we collected a separate prospective cohort of 114 samples to validate the discriminative power of the trained prediction models. Using machine learning under cross validation setting, the feature selection method selected a concise biomarker panel for each CKD-MBD subtype as well as for the general one. Using the consensus of these features, best area under ROC curve reached up to 0.95 for the training dataset and 0.74 for the perspective dataset, respectively. For the first time, we utilized machine learning methods to analyze biochemical criteria associated with CKD-MBD. Our aim was to identify alternative biomarkers that could serve not only as early detection indicators for CKD-MBD, but also as potential candidates for studying the underlying molecular mechanisms of the condition.

Abnormal Heart Sound Detection using Time-Frequency Analysis and Machine Learning Techniques

Phonocardiogram (PCG) signals contains valuable information pertaining to heart valve functionality, rendering them potentially useful for early detection of cardiovascular diseases. Automated classification of heart sounds harbors great promise for identifying cardiac pathologies. This paper introduces a novel automated approach to classify normal and abnormal heart sounds. Our methodology involves partitioning heart sounds into four segments: S1, S2, systolic, and diastolic, followed by extraction of time–frequency and time-statistical features. Prior to data classification, we employ two techniques - particle Swarm optimization (PSO) and Sequential Forward Feature Selection (SFFS) - for efficient feature selection. We assess the performance of the proposed method on the Physionet Challenge 2016 database, utilizing the 10-fold cross-validation method. To address the issue of dataset imbalance, we apply the synthetic minority over-sampling technique (SMOTE) to create balanced datasets. Our approach surpasses existing methods in the literature, as evidenced by its superior accuracy, sensitivity, and specificity metrics. Specifically, our method achieves an accuracy of 98.03%, a sensitivity of 97.64%, and a specificity of 98.43% in distinguishing normal from abnormal heart sounds on the Physionet database. These findings outperform the results obtained by previously established methods evaluated on the Physionet 2016 challenge database.

Personalized Feature Importance Ranking for Affect Recognition From Behavioral and Physiological Data

Designing affect-based personalized technology involves dealing with large datasets. Machine learning (ML) algorithms are employed to predict affect and similar human factors from in-game metrics, behavioral patterns, or physiological responses. The classification performance is usually presented as a global point estimate without providing user-specific interpretations. This approach is incompatible with effective personalization in games because it disregards the variability of body responses between players. This paper proposes a methodology to classify subjective human factors from large multimodal data. A public VR dataset (CEAP-360VR) was used to extensively compare three ML classifiers and five feature importance techniques. The produced models could reduce the original feature space by 82% (from 113 to 20 features) without compromising predictive performance (F1 score). A random forest (RF) using forward sequential feature selection (fSFS) yielded the best prediction of binary valence (F1=0.761) and arousal (F1=0.748). Finally, feature importance rankings are discussed with emphasis on global and user-specific patterns that may improve affect recognition. The proposed methodology is envisioned to help game designers and researchers create customized user-centric games and VR experiences inferring possible explanations from multimodal datasets.

Prediction Model of Ocular Metastases in Gastric Adenocarcinoma: Machine Learning-Based Development and Interpretation Study.

Background: Although gastric adenocarcinoma (GA) related ocular metastasis (OM) is rare, its occurrence indicates a more severe disease. We aimed to utilize machine learning (ML) to analyze the risk factors of GA-related OM and predict its risks. Methods: This is a retrospective cohort study. The clinical data of 3532 GA patients were collected and randomly classified into training and validation sets in a ratio of 7:3. Those with or without OM were classified into OM and non-OM (NOM) groups. Univariate and multivariate logistic regression analyses and least absolute shrinkage and selection operator were conducted. We integrated the variables identified through feature importance ranking and further refined the selection process using forward sequential feature selection based on random forest (RF) algorithm before incorporating them into the ML model. We applied six ML algorithms to construct the predictive GA model. The area under the receiver operating characteristic (ROC) curve indicated the model's predictive ability. Also, we established a network risk calculator based on the best performance model. We used Shapley additive interpretation (SHAP) to identify risk factors and to confirm the interpretability of the black box model. We have de-identified all patient details. Results: The ML model, consisting of 13 variables, achieved an optimal predictive performance using the gradient boosting machine (GBM) model, with an impressive area under the curve (AUC) of 0.997 in the test set. Utilizing the SHAP method, we identified crucial factors for OM in GA patients, including LDL, CA724, CEA, AFP, CA125, Hb, CA153, and Ca2+. Additionally, we validated the model's reliability through an analysis of two patient cases and developed a functional online web prediction calculator based on the GBM model. Conclusion: We used the ML method to establish a risk prediction model for GA-related OM and showed that GBM performed best among the six ML models. The model may identify patients with GA-related OM to provide early and timely treatment.

Wind power plants hybridised with solar power: A generation forecast perspective

The association of different variable renewable technologies in hybrid power plants and the benefits of their aggregation for the operation of power systems is an area of recent research. Accurate forecasts are crucial for extracting those benefits and promote an optimal integration of such plants into power systems and electricity markets. This study focuses on the hybridisation of existing wind power plants with different shares of solar photovoltaic capacity and investigates how these power plants can reduce their combined forecast errors and thus, increasing profitability in electricity markets.The work uses a forecast methodology based on a sequential forward feature selection algorithm which employs two different objective functions and an artificial neural network approach previously presented but, in this case, it is applied to the specific case of hybrid power plants. The methodology uses as input data from a numerical weather prediction model and iteratively selects meteorological features to achieve the different objective functions implemented, namely i) minimisation of the root mean square error; or ii) maximisation of the market remuneration.The methodology developed was applied to three case studies in Portugal with different levels of wind and solar generation complementarity. The results show that the hybrid power plants can increase market value by up to 5% and total remuneration can increase by up to 30% when compared with the existing wind power plant, while it is possible to reduce the forecast errors by nearly 4%. The obtained results highlight the need to select the most relevant meteorological features to maximise the accuracy of the power forecast and the renewable power producers revenues in a market environment.

Deep Learning Based Hybrid Intrusion Detection Systems to Protect Satellite Networks

Despite the fact that satellite-terrestrial systems have advantages such as high throughput, low latency, and low energy consumption, as well as low exposure to physical threats and natural disasters and cost-effective global coverage, their integration exposes both of them to particular security challenges that can arise due to the migration of security challenges from one to another. Intrusion Detection Systems (IDS) can also be used to provide a high level of protection for modern network environments such as satellite-terrestrial integrated networks (STINs). To optimize the detection performance of malicious activities in network traffic, four hybrid intrusion detection systems for satellite-terrestrial communication systems (SAT-IDSs) are proposed in this paper. All the proposed systems exploit the sequential forward feature selection (SFS) method based on random forest (RF) to select important features from the dataset that increase relevance and reduce complexity and then combine them with a machine learning (ML) or deep learning (DL) model; Random Forest (RF), Long Short-Term memory (LSTM), Artificial Neural Networks (ANN), and Gated Recurrent Unit (GRU). Two datasets—STIN, which simulates satellite networks, and UNSW-NB15, which simulates terrestrial networks—were used to evaluate the performance of the proposed SAT-IDSs. The experimental results indicate that selecting significant and crucial features produced by RF-SFS vastly improves detection accuracy and computational efficiency. In the first dataset (STIN), the proposed hybrid ML system SFS-RF achieved an accuracy of 90.5% after using 10 selected features, compared to 85.41% when using the whole dataset. Furthermore, the RF-SFS-GRU model achieved the highest performance of the three proposed hybrid DL-based SAT-IDS with an accuracy of 87% after using 10 selected features, compared to 79% when using the entire dataset. In the second dataset (UNSW-NB15), the proposed hybrid ML system SFS-RF achieved an accuracy of 78.52% after using 10 selected features, compared to 75.4% when using the whole dataset. The model with the highest accuracy of the three proposed hybrid DL-based SAT-IDS was the RF-SFS-GRU model. It achieved an accuracy of 79% after using 10 selected features, compared to 74% when using the whole dataset.

Using Single-Voxel Magnetic Resonance Spectroscopy Data Acquired at 1.5T to Classify Multivoxel Data at 3T: A Proof-of-Concept Study.

To test whether MV grids can be classified with models trained with SV. Retrospective study. Training dataset: Multicenter multiformat SV INTERPRET, 1.5T. Testing dataset: MV eTumour, 3T. Two classification tasks were completed: 3-class (meningioma vs. aggressive vs. normal) and 4-class (meningioma vs. low-grade glioma vs. aggressive vs. normal). Five different methods were tested for feature selection. The classification was implemented using linear discriminant analysis (LDA), random forest, and support vector machines. The evaluation was completed with balanced error rate (BER) and area under the curve (AUC) on both sets. The accuracy in class prediction was calculated by developing a solid tumor index (STI) and segmentation accuracy with the Dice score. The best method was sequential forward feature selection combined with LDA, with AUCs = 0.95 (meningioma), 0.89 (aggressive), 0.82 (low-grade glioma), and 0.82 (normal). STI was 66% (4-class task) and 71% (3-class task) because two cases failed completely and two more had suboptimal STI as defined by us. The reasons for failure in the classification of the MV test set were related to the presence of artifacts.

SFRP1 Expression is Inversely Associated With Metastasis Formation in Canine Mammary Tumours

BackgroundCanine mammary tumours (CMTs) are the most frequent tumours in intact female dogs and show strong similarities with human breast cancer. In contrast to the human disease there are no standardised diagnostic or prognostic biomarkers available to guide treatment. We recently identified a prognostic 18-gene RNA signature that could stratify human breast cancer patients into groups with significantly different risk of distant metastasis formation. Here, we assessed whether expression patterns of these RNAs were also associated with canine tumour progression.MethodA sequential forward feature selection process was performed on a previously published microarray dataset of 27 CMTs with and without lymph node (LN) metastases to identify RNAs with significantly differential expression to identify prognostic genes within the 18-gene signature. Using an independent set of 33 newly identified archival CMTs, we compared expression of the identified prognostic subset on RNA and protein basis using RT-qPCR and immunohistochemistry on FFPE-tissue sections.ResultsWhile the 18-gene signature as a whole did not have any prognostic power, a subset of three RNAs: Col13a1, Spock2, and Sfrp1, together completely separated CMTs with and without LN metastasis in the microarray set. However, in the new independent set assessed by RT-qPCR, only the Wnt-antagonist Sfrp1 showed significantly increased mRNA abundance in CMTs without LN metastases on its own (p = 0.013) in logistic regression analysis. This correlated with stronger SFRP1 protein staining intensity of the myoepithelium and/or stroma (p < 0.001). SFRP1 staining, as well as β-catenin membrane staining, was significantly associated with negative LN status (p = 0.010 and 0.014 respectively). However, SFRP1 did not correlate with β-catenin membrane staining (p = 0.14).ConclusionThe study identified SFRP1 as a potential biomarker for metastasis formation in CMTs, but lack of SFRP1 was not associated with reduced membrane-localisation of β-catenin in CMTs.

A novel machine learning-based feature extraction method for classifying intracranial hemorrhage computed tomography images

One of the most serious forms of brain stroke is intracranial hemorrhage (ICH). When an artery bursts, the brain and the tissue around the artery start bleeding. This study proposes a joint feature selection strategy to classify computed tomography (CT) images of intracranial hemorrhage. The joint feature set is composed of transform and texture features. Joint features are constructed from a combination of grey level co-occurrence matrix (GLCM) features, discrete wavelet features (DWT), and discrete cosine features (DCT). Brain hemorrhage CT image classification uses ensemble-based machine learning (ML) techniques. On the training dataset, a Synthetic Minority Over-Sampling Technique (SMOTE) is applied to treat the problem of oversampling by adding fresh data. Additionally, the sequential forward feature selection technique is used to obtain feature subsets. The classification accuracy is further examined for varied feature vector sizes. Confusion matrix, precision, and recall in categorization are employed as performance evaluation measurements. The ML-based ensemble classifiers can produce highly accurate results with the aid of the proposed novel feature extraction mechanism. When taking into consideration a crucial feature set consisting of six features, it can be seen that Random Forest obtained the greatest accuracy, which is 87.22%.

Predicting Football Team Performance with Explainable AI: Leveraging SHAP to Identify Key Team-Level Performance Metrics

Understanding the performance indicators that contribute to the final score of a football match is crucial for directing the training process towards specific goals. This paper presents a pipeline for identifying key team-level performance variables in football using explainable ML techniques. The input data includes various team-specific features such as ball possession and pass behaviors, with the target output being the average scoring performance of each team over a season. The pipeline includes data preprocessing, sequential forward feature selection, model training, prediction, and explainability using SHapley Additive exPlanations (SHAP). Results show that 14 variables have the greatest contribution to the outcome of a match, with 12 having a positive effect and 2 having a negative effect. The study also identified the importance of certain performance indicators, such as shots, chances, passing, and ball possession, to the final score. This pipeline provides valuable insights for coaches and sports analysts to understand which aspects of a team’s performance need improvement and enable targeted interventions to improve performance. The use of explainable ML techniques allows for a deeper understanding of the factors contributing to the predicted average team score performance.

PO-02-021 DIRECTION OF REPOLARIZATION IN BASELINE ECG IS PREDICTIVE OF RESPONSE TO CARDIAC RESYNCHRONIZATION THERAPY

Cardiac resynchronization therapy (CRT) is a cornerstone of treatment for advanced heart failure, but not all patients derive equal benefit from this intervention. Current approaches of using electrocardiographic and echocardiographic parameters to identify likely CRT responders demonstrate poor predictive ability. We seek to identify electrocardiographic parameters that can improve prediction of benefit with CRT This was a retrospective analysis of patients who underwent CRT-D placement at University of Pittsburgh Medical Center from 2014-2020. Patients were excluded if pre-implantation EKG or pre- and post-implantation echocardiograms were not available. Each patient’s immediate pre-implantation ECG was processed using the CineECG software (EKG Excellence) to obtain the mean temporospatial isochrone of ventricular activation. 192 features quantifying the activation and recovery pathways were computed for each patient. Response to CRT was defined as an improvement in ejection fraction (EF) of 5% on follow-up transthoracic echocardiogram 3-12 months after CRT implantation. Forward sequential feature selection was used to identify the features of highest importance based on negative logistic loss. 276 patients were included in this analysis (36% female, mean age 69 years, mean pre-procedural EF 27%). The 10 most important CineECG features were identified. When these features were used in a multivariate logistic regression, three features were independently predictive of CRT response: left bundle branch block (LBBB); leftward direction of late repolarization potentials (i.e., Tpeak - Tend rotated toward LV in the horizontal plane); and location of center-of-mass of terminal cardiac activation. Compared to CRT responders (n=169, 62%), non-responders were more likely to have abnormal direction of late repolarization potentials in both LBBB group (-0.34±0.30 vs. -0.45±0.30 vs, p=0.027) and RBBB group (0.25±0.51 vs. 0.61±0.27, p=0.031). In the latter group, direction of later repolarization potentials correlated with magnitude of change in post-procedural EF (r=0.23, p=0.016). Our findings confirm that the presence of LBBB is predictive of response to CRT. In addition, our data suggests that in patients with RBBB, a more leftward direction of late repolarization potentials is independently predictive of CRT response. Further studies into repolarization dynamics as a predictor of CRT response in RBBB are warranted.

Predicting perinatal mortality based on maternal health status and health insurance service using homogeneous ensemble machine learning methods

BackgroundPerinatal mortality in Ethiopia is the highest in Africa, with 68 per 1000 pregnancies intrapartum deaths. It is mainly associated with home delivery, which contributes to more than 75% of perinatal deaths. Financial constraints significantly impact timely access to maternal health care. Financial incentives, such as health insurance, may address the demand- and supply-side factors. This study, hence, aims to predict perinatal mortality based on maternal health status and health insurance service using homogeneous ensemble machine learning methods.MethodsThe data was collected from the Ethiopian demographic health survey from 2011 to 2019 G.C. The data were pre-processed to get quality data that are suitable for the homogenous ensemble machine-learning algorithms to develop a model that predicts perinatal mortality. We have applied filter (chi-square and mutual information) and wrapper (sequential forward and sequential backward) feature selection methods. After selecting all the relevant features, we developed a predictive model using cat boost, random forest, and gradient boosting algorithms and evaluated the model using both objective (accuracy, precision, recall, F1_score, ROC) and subjective (domain expert) based evaluation techniques.ResultsPerinatal mortality prediction models were developed using random forest, gradient boosting, and cat boost algorithms with the overall accuracy of 89.95%, 90.24%, and 82%, respectively. Risk factors of perinatal mortality were identified using feature importance analysis and relevant rules were extracted using the best performing model.ConclusionsA prediction model that was developed using gradient boosting algorithms was selected for further use in the risk factor analysis, generating relevant rules, development of artifacts, and model deployment because it has registered better performance with 90.24% accuracy. The most determinant risk factors of perinatal mortality were identified using feature importance and some of them are community-based health insurance, mother's educational level, region and place of residence, age, wealth status, birth interval, preterm, smoking cigarette, anemia level, hemoglobin level, and marital status.

Correlation of image textures of a polarization feature parameter and the microstructures of liver fibrosis tissues

Mueller matrix imaging is emerging for the quantitative characterization of pathological microstructures and is especially sensitive to fibrous structures. Liver fibrosis is a characteristic of many types of chronic liver diseases. The clinical diagnosis of liver fibrosis requires time-consuming multiple staining processes that specifically target on fibrous structures. The staining proficiency of technicians and the subjective visualization of pathologists may bring inconsistency to clinical diagnosis. Mueller matrix imaging can reduce the multiple staining processes and provide quantitative diagnostic indicators to characterize liver fibrosis tissues. In this study, a fiber-sensitive polarization feature parameter (PFP) was derived through the forward sequential feature selection (SFS) and linear discriminant analysis (LDA) to target on the identification of fibrous structures. Then, the Pearson correlation coefficients and the statistical T-tests between the fiber-sensitive PFP image textures and the liver fibrosis tissues were calculated. The results show the gray level run length matrix (GLRLM)-based run entropy that measures the heterogeneity of the PFP image was most correlated to the changes of liver fibrosis tissues at four stages with a Pearson correlation of 0.6919. The results also indicate the highest Pearson correlation of 0.9996 was achieved through the linear regression predictions of the combination of the PFP image textures. This study demonstrates the potential of deriving a fiber-sensitive PFP to reduce the multiple staining process and provide textures-based quantitative diagnostic indicators for the staging of liver fibrosis.

Enhancing wind power forecast accuracy using the weather research and forecasting numerical model-based features and artificial neuronal networks

Forecasting with accuracy the quantity of energy produced by wind power plants is crucial to enabling its optimal integration into power systems and electricity markets. Despite the remarkable improvements in the wind forecasting systems in recent years, large errors can still be observed, especially for longer time horizons. This work focuses on identifying new numerical weather prediction (NWP)-based features aiming to improve the overall quality of wind power forecasts.The methodology also incorporates a sequential forward feature selection algorithm. This algorithm was designed to select iteratively the meteorological features which minimize the wind forecast errors.The methodology was applied separately to seven wind parks in Portugal with different climate characteristics. The proposed approach allowed a reduction between 13% and 37% in the root mean square errors of wind power forecasts, compared with a baseline scenario. While the meteorological features identified for each wind park showed similarities within regions with analogous wind power generation profiles, each wind park required specific meteorological parameters as input data to obtain the best performance. Thus, the results show to be crucial to select the most relevant features of a specific site to maximize the accuracy of a wind power forecast.

Fully automated radiomics-based machine learning models for multiclass classification of single brain tumors: Glioblastoma, lymphoma, and metastasis

Background and purposeTo investigate the diagnostic performance of fully automated radiomics-based models for multiclass classification of a single enhancing brain tumor among glioblastoma, central nervous system lymphoma, and metastasis. Materials and methodsThe training and test sets were comprised of 538 cases (300 glioblastomas, 73 lymphomas, and 165 metastases) and 169 cases (101 glioblastomas, 29 lymphomas, and 39 metastases), respectively. After fully automated segmentation, radiomic features were extracted. Three conventional machine learning classifiers, including least absolute shrinkage and selection operator (LASSO), adaptive boosting (Adaboost), and support vector machine with the linear kernel (SVC), combined with one of four feature selection methods, including forward sequential feature selection, F score, mutual information, and LASSO, were trained. Additionally, one ensemble classifier based on the three classifiers was used. The diagnostic performance of the optimized models was tested in the test set using the accuracy, F1-macro score, and the area under the receiver operating characteristic curve (AUCROC). ResultsThe best performance was achieved when the LASSO was used as a feature selection method. In the test set, the best performance was achieved by the ensemble classifier, showing an accuracy of 76.3% (95% CI, 70.0–82.7), a F1-macro score of 0.704, and an AUCROC of 0.878. ConclusionOur fully automated radiomics-based models for multiclass classification might be useful for differential diagnosis of a single enhancing brain tumor with a good diagnostic performance and generalizability.

Load2Load: Day-ahead load forecasting at aggregated level

A reliable and accurate short-term load forecasting (STLF) helps utilities and energy providers deal with the challenges posed by supply and demand balance, higher penetration of renewable energies and the development of electricity markets with increasingly complex pricing strategies in future smart grids. Recent advances in deep learning have been successively utilized to STLF. However, there is no certain study that evaluates the performances of different STLF methods at an aggregated level on different datasets with different numbers of daily measurements.In this study, a deep learning STLF architecture called Load2Load is proposed for day-ahead forecasting. Different forecasting methods have been evaluated and compared on two datasets with different temporal resolutions and features. An additive ensemble method as well as a selective ensemble method that selects the outputs of different forecasters in an hourly manner are proposed. Moreover; a modified sequential forward feature selection algorithm is proposed, resulting in better performance with a much smaller number of features.Numerical results show that the proposed Load2Load architecture has a competing performance compared to other advanced forecasters. When used together with the proposed ensemble methods, Load2Load can significantly improve the forecasting performance. The proposed feature selection algorithm results in better performance for the majority of the cases while reducing the dimensionality. According to the results with two different datasets; the proposed methods are shown to be robust to temporal resolutions, feature types and sequence lengths.

Comparison between handcraft feature extraction and methods based on Recurrent Neural Network models for gesture recognition by instrumented gloves: A case for Brazilian Sign Language Alphabet

Sign language is the communication mode used by hearing-impaired people. The development of tools for recognition of sign language is important to improve the interaction of these people in the society. This work proposes a comparison between models of gesture recognition based on instrumented glove with and without handcrafted feature extraction approach, which is a classical method for pattern recognition. For manual feature extraction approach, Forward Sequential Feature Selection (SFS) method was employed to select the best features and different classifiers were analyzed. For the non-manual feature extraction approach, Recurrent Neural Networks as Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) models and, additionally, a proposed Ensemble GRU model were employed. The proposed Ensemble GRU is composed of three layers, and each layer is responsible to acquire knowledge of gestures’ periods: initial, gesture, and final periods. The classifiers LSTM and GRU can receive the data segments of the gestures directly, without manual feature extraction. A dataset composed by 26 gestures from alphabet of Brazilian Sign Language was explored. In the handcrafted feature extraction approach, the best result was of 88.5% accuracy, obtained with K-Nearest Neighbors model. On the opposite model, GRU and Ensemble GRU models reached 94.6% and 96.8% accuracy. It can be observed that non-manual feature extraction approach can be implemented for gesture recognition based on instrumented glove and the results can be better than conventional methods. Furthermore, the proposed Ensemble GRU model presented better result than the standard models of GRU and LSTM.