Heart disease, one of the main reasons behind the high mortality rate around the world, requires a sophisticated and expensive diagnosis process. In the recent past, much literature has demonstrated machine learning approaches as an opportunity to efficiently diagnose heart disease patients. However, challenges associated with datasets such as missing data, inconsistent data, and mixed data (containing inconsistent missing data both as numerical and categorical) are often obstacles in medical diagnosis. This inconsistency led to a higher probability of misprediction and a misled result. Data preprocessing steps like feature reduction, data conversion, and data scaling are employed to form a standard dataset—such measures play a crucial role in reducing inaccuracy in final prediction. This paper aims to evaluate eleven machine learning (ML) algorithms—Logistic Regression (LR), Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Classification and Regression Trees (CART), Naive Bayes (NB), Support Vector Machine (SVM), XGBoost (XGB), Random Forest Classifier (RF), Gradient Boost (GB), AdaBoost (AB), Extra Tree Classifier (ET)—and six different data scaling methods—Normalization (NR), Standscale (SS), MinMax (MM), MaxAbs (MA), Robust Scaler (RS), and Quantile Transformer (QT) on a dataset comprising of information of patients with heart disease. The result shows that CART, along with RS or QT, outperforms all other ML algorithms with 100% accuracy, 100% precision, 99% recall, and 100% F1 score. The study outcomes demonstrate that the model’s performance varies depending on the data scaling method.

No abstract

The extraction methodology of the significant features from the signals is one of the most important prerequisite steps for EEG signal classification. Common spatial pattern (CSP) is a widely used feature extraction method for EEG signal but with a lacking of failing to maintain discriminative features between classes in the time domain, and further as a consequence, ends up in inconvenience with erroneous output. To overcome the limitations of the convention CSP, this research work proposes a novel frequency domain CSP (FCSP) method for feature extraction. This method proposes to convert the time domain EEG signal to its power spectral density (PSD) so that the event-related variation can be found in the frequency domain. After that, the CSP method is applied to the PSD values of the selected channels to extract the variation based on the spatial pattern of the channels for the events. The output of this method helps to extract simple features from the FCSP-PSD data for the classification. The proposed method is applied to motor imagery data from BCI competition IV. To check the applicability of the proposed method, a complex environment was created considering the same lobe events such as combined left and right feet (Class#1) versus right-hand (Class#2) imagery movement. To compare the performance of the proposed work, the method is also applied to the conventional classification problem (left-hand vs right-hand imagery movement) and found very promising results of 91% accuracy on average.

No abstract