Random Forest
Recently Published Documents





2021 ◽  
Vol 2 ◽  
Cornelia Sindermann ◽  
René Mõttus ◽  
Dmitri Rozgonjuk ◽  
Christian Montag

To understand what was driving individual differences in voting intentions in a large German sample, we investigated the predictability of voting intentions from the Big Five personality domains, facets, and nuances, thereby tackling shortcomings of previous studies. Using random forest analyses in a dataset of N = 4,286 individuals (46.01% men), separate models were trained to predict intentions to 1) not vote versus to vote, 2) vote for a specific party, and 3) vote for a left- versus right-from-the-center party from either the Big Five personality domains, facets, or nuances (represented by individual items). Except for intentions to not vote versus to vote, balanced accuracies to predict voting intentions marginally exceeded those achieved by a baseline learner always predicting the majority class. Using nuances over facets and domains slightly increased balanced accuracies. Results indicate that additional variables should be considered to accurately predict voting intentions, at least in German samples.

Georgi P. Dimitrov ◽  
Galina Panayotova ◽  
Boyan Jekov ◽  
Pavel Petrov ◽  
Iva Kostadinova ◽  

Comparison of the Accuracy of different off-line methods for classification Electroencephalograph (EEG) signals, obtained from Brain-Computer Interface (BCI) devices are investigated in this paper. BCI is a technology that allows people to interact directly or indirectly with their environment only by using brain activity. But, the method of signal acquisition is non-invasive, resulting in significant data loss. In addition, the received signals do not contain only useful information. All this requires careful selection of the method for the classification of the received signals. The main purpose of this paper is to provide a fair and extensive comparison of some commonly employed classification methods under the same conditions so that the assessment of different classifiers will be more convictive. In this study, we investigated the accuracy of the classification of the received signals with classifiers based on AdaBoost (AB), Decision Tree (DT), k-Nearest Neighbor (kNN), Gaussian SVM, Linear SVM, Polynomial SVM, Random Forest (RF), Random Forest Regression ( RFR ). We used only basic parameters in the classification, and we did not apply fine optimization of the classification results. The obtained results show suitable algorithms for the classification of EEG signals. This would help young researchers to achieve interesting results in this field faster.

2021 ◽  
Vol 12 ◽  
Jernej Rosker ◽  
Ziva Majcen Rosker

Analyzing visual search strategies in tennis is primarily focused on studying relationships between visual behavior and tennis performance. However, diverse movement characteristics among different servers suggest the importance of adjusting the visual search strategies of an individual while playing against different opponents. The aim of this study was to analyze whether visual search strategies can be attributed to the individual server and the returning player during the tennis serve return or return performance. Seventeen tennis players were enrolled in this study (five international players and 12 national players) producing a sample of 1,020 returns measured with mobile eye trackers. The random forest machine learning model was used to analyze the ability to classify the returning player [area under the curve (AUC): 0.953], individual server (AUC: 0.686), and return performance category (AUC: 0.667) based on the location and duration of the focal vision fixation. In international tennis players, the higher predictability of the server was observed as compared with national level players (AUC: 0.901 and 0.834, respectively). More experienced tennis players presented with a higher ability to adjust their visual search strategies to different servers. International players also demonstrated anticipatory visual behavior during the tossing hand movement and superior information pickup during the final phases of the stroke of a server.

Weiheng Wen ◽  
Peili Wu ◽  
Yugang Zhang ◽  
Zijian Chen ◽  
Jia Sun ◽  

Objective: Non-alcoholic fatty liver disease (NAFLD) is a serious health threat worldwide. The aim of this study was to comprehensively describe the metabolic and immunologic characteristics of NAFLD, and to explore potential therapeutic drug targets for NAFLD.Methods: Six NAFLD datasets were downloaded from the Gene Expression Omnibus (GEO) database, including GSE48452, GSE63067, GSE66676, GSE89632, GSE24807, and GSE37031. The datasets we then used to identify and analyze genes that were differentially expressed in samples from patients with NAFLD and normal subjects, followed by analysis of the metabolic and immunologic characteristics of patients with NAFLD. We also identified potential therapeutic drugs for NAFLD using the Connectivity Map (CMAP) database. Moreover, we constructed a prediction model using minimum depth random forest analysis and screened for potential therapeutic targets. Finally, therapeutic targets were verified in a fatty liver model stimulated by palmitic acid (PA).Results: A total of 1,358 differentially expressed genes (DEGs) were obtained, which were mainly enriched in carbohydrate metabolism, lipid metabolism, and other metabolic pathways. Immune infiltration analysis showed that memory B cells, regulatory T cells and M1 macrophage were significantly up-regulated, while T cells follicular helper were down regulated in NAFLD. These may provide a reference for the immune-metabolism interaction in the pathogenesis of NAFLD. Digoxin and helveticoside were identified as potential therapeutic drugs for NAFLD via the CMAP database. In addition, a five-gene prediction model based on minimum depth random forest analysis was constructed, and the receiver operating characteristic (ROC) curves of both training and validation set reached 1. The five candidate therapeutic targets were ENO3, CXCL10, INHBE, LRRC31, and OPTN. Moreover, the efficiency of hepatocyte adipogenesis decreased after OPTN knockout, confirming the potential use of OPTN as a new therapeutic target for NAFLD.Conclusion: This study provides a deeper insight into the molecular pathogenesis of NAFLD. We used five key genes to construct a diagnostic model with a strong predictive effect. Therefore, these five key genes may play an important role in the diagnosis and treatment of NAFLD, particularly those with increased OPTN expression.

2021 ◽  
Vol 11 (1) ◽  
Eiman Alothali ◽  
Kadhim Hayawi ◽  
Hany Alashwal

AbstractThe last few years have revealed that social bots in social networks have become more sophisticated in design as they adapt their features to avoid detection systems. The deceptive nature of bots to mimic human users is due to the advancement of artificial intelligence and chatbots, where these bots learn and adjust very quickly. Therefore, finding the optimal features needed to detect them is an area for further investigation. In this paper, we propose a hybrid feature selection (FS) method to evaluate profile metadata features to find these optimal features, which are evaluated using random forest, naïve Bayes, support vector machines, and neural networks. We found that the cross-validation attribute evaluation performance was the best when compared to other FS methods. Our results show that the random forest classifier with six optimal features achieved the best score of 94.3% for the area under the curve. The results maintained overall 89% accuracy, 83.8% precision, and 83.3% recall for the bot class. We found that using four features: favorites_count, verified, statuses_count, and average_tweets_per_day, achieves good performance metrics for bot detection (84.1% precision, 81.2% recall).

Sign in / Sign up

Export Citation Format

Share Document