Feature Selection Methods Simultaneously Improve the Detection Accuracy and Model Building Time of Machine Learning Classifiers

Corticosteroid sensitivity detection in sepsis patients using a personalized data mining approach: A clinical investigation

BACKGROUND AND OBJECTIVE

Sepsis is a life-threatening disease with high mortality, incidence, and morbidity. Corticosteroids (CS) are a recommended treatment for sepsis, but some patients respond negatively to CS therapy. Early prediction of corticosteroid responsiveness can help intervene and reduce mortality. In this study, we aim to develop a data mining methodology for predicting CS responsiveness of septic patients.

METHODS

We used data from a randomized controlled trial called APROCCHSS, which recruited 1241 septis patients to study the effectiveness of corticotherapy. We conducted a thorough study of multiple machine learning models to select the most efficient prediction model, called "signature". We evaluated the performance of the signature using precision, sensitivity, and specificity values.

RESULTS

We found that Logistic Regression was the best model with an AUC of 72%. We conducted further experiments to examine the impact of additional features and the model's generalizability to different groups of patients. We also performed a statistical analysis to analyze the effect of the treatment at the individual level and on the population as a whole.

CONCLUSIONS

Our data mining methodology can accurately predict cortico-sensitivity or resistance in septis patients. The signature has been deployed into the Assistance Publique - Hôpitaux de Paris (APHP) information system as a web service, taking patient information as input and providing a prediction of cortico-sensitivity or resistance. Early prediction of corticosteroid responsiveness can help clinicians intervene promptly and improve patient outcomes.

Explainable machine learning models based on multimodal time-series data for the early detection of Parkinson's disease

BACKGROUND AND OBJECTIVES

Parkinson's Disease (PD) is a devastating chronic neurological condition. Machine learning (ML) techniques have been used in the early prediction of PD progression. Fusion of heterogeneous data modalities proved its capability to improve the performance of ML models. Time series data fusion supports the tracking of the disease over time. In addition, the trustworthiness of the resulting models is improved by adding model explainability features. The literature on PD has not sufficiently explored these three points.

METHODS

In this work, we proposed an ML pipeline for predicting the progression of PD that is both accurate and explainable. We explore the fusion of different combinations of five time series modalities from the Parkinson's Progression Markers Initiative (PPMI) real-world dataset, including patient characteristics, biosamples, medication history, motor, and non-motor function data. Each patient has six visits. The problem has been formulated in two ways: ❶ a three-class based progression prediction with 953 patients in each time series modality, and ❷ a four-class based progression prediction with 1,060 patients in each time series modality. The statistical features of these six visits were calculated from each modality and diverse feature selection methods were applied to select the most informative feature sets. The extracted features were used to train a set of well-known ML models including Support vector machines (SVM), random forests (RF), extra tree classifier (ETC), light gradient boosting machines (LGBM), and stochastic gradient descent (SGD). We examined a number of data-balancing strategies in the pipeline with different combinations of modalities. ML models have been optimized using the Bayesian optimizer. A comprehensive evaluation of various ML methods has been conducted, and the best models have been extended to provide different explainability features.

RESULTS

We compare the performance of ML models before and after optimization and using and without using feature selection. In the three-class experiment and with various modality fusions, the LGBM model produced the most accurate results with a 10-fold cross-validation (10-CV) accuracy of 90.73% using non-motor function modality. RF produced the best results in the four-class experiment with various modality fusions with a 10-CV accuracy of 94.57% using non-motor modality. With the fused dataset of non-motor and motor function modalities, the LGBM model outperformed the other ML models in both the 3-class and 4-class experiments (i.e., 10-CV accuracy of 94.89% and 93.73%, respectively). Using the Shapely Additive Explanations (SHAP) framework, we employed global and instance-based explanations to explain the behavior of each ML classifier. Moreover, we extended the explainability by implementing the LIME and SHAPASH local explainers. The consistency of these explainers has been explored. The resultant classifiers were accurate, explainable, and thus medically more relevant and applicable.

CONCLUSIONS

The select modalities and feature sets were confirmed by the literature and medical experts. The various explainers suggest that the bradykinesia (NP3BRADY) feature was the most dominant and consistent. By providing thorough insights into the influence of multiple modalities on the disease risk, the suggested approach is expected to help improve the clinical knowledge of PD progression processes.

An Anomaly Intrusion Detection for High-Density Internet of Things Wireless Communication Network Based Deep Learning Algorithms

Telecommunication networks are growing exponentially due to their significant role in civilization and industry. As a result of this very significant role, diverse applications have been appeared, which require secured links for data transmission. However, Internet-of-Things (IoT) devices are a substantial field that utilizes the wireless communication infrastructure. However, the IoT, besides the diversity of communications, are more vulnerable to attacks due to the physical distribution in real world. Attackers may prevent the services from running or even forward all of the critical data across the network. That is, an Intrusion Detection System (IDS) has to be integrated into the communication networks. In the literature, there are numerous methodologies to implement the IDSs. In this paper, two distinct models are proposed. In the first model, a custom Convolutional Neural Network (CNN) was constructed and combined with Long Short Term Memory (LSTM) deep network layers. The second model was built about the all fully connected layers (dense layers) to construct an Artificial Neural Network (ANN). Thus, the second model, which is a custom of an ANN layers with various dimensions, is proposed. Results were outstanding a compared to the Logistic Regression algorithm (LR), where an accuracy of 97.01% was obtained in the second model and 96.08% in the first model, compared to the LR algorithm, which showed an accuracy of 92.8%.

An Efficient Hyperparameter Control Method for a Network Intrusion Detection System Based on Proximal Policy Optimization

The complexity of network intrusion detection systems (IDSs) is increasing due to the continuous increases in network traffic, various attacks and the ever-changing network environment. In addition, network traffic is asymmetric with few attack data, but the attack data are so complex that it is difficult to detect one. Many studies on improving intrusion detection performance using feature engineering have been conducted. These studies work well in the dataset environment; however, it is challenging to cope with a changing network environment. This paper proposes an intrusion detection hyperparameter control system (IDHCS) that controls and trains a deep neural network (DNN) feature extractor and k-means clustering module as a reinforcement learning model based on proximal policy optimization (PPO). An IDHCS controls the DNN feature extractor to extract the most valuable features in the network environment, and identifies intrusion through k-means clustering. Through iterative learning using the PPO-based reinforcement learning model, the system is optimized to improve performance automatically according to the network environment, where the IDHCS is used. Experiments were conducted to evaluate the system performance using the CICIDS2017 and UNSW-NB15 datasets. In CICIDS2017, an F1-score of 0.96552 was achieved and UNSW-NB15 achieved an F1-score of 0.94268. An experiment was conducted by merging the two datasets to build a more extensive and complex test environment. By merging datasets, the attack types in the experiment became more diverse and their patterns became more complex. An F1-score of 0.93567 was achieved in the merged dataset, indicating 97% to 99% performance compared with CICIDS2017 and UNSW-NB15. The results reveal that the proposed IDHCS improved the performance of the IDS by automating learning new types of attacks by managing intrusion detection features regardless of the network environment changes through continuous learning.

Applied Identification of Industry Data Science Using an Advanced Multi-Componential Discretization Model

Applied human large-scale data are collected from heterogeneous science or industry databases for the purposes of achieving data utilization in complex application environments, such as in financial applications. This has posed great opportunities and challenges to all kinds of scientific data researchers. Thus, finding an intelligent hybrid model that solves financial application problems of the stock market is an important issue for financial analysts. In practice, classification applications that focus on the earnings per share (EPS) with financial ratios from an industry database often demonstrate that the data meet the abovementioned standards and have particularly high application value. This study proposes several advanced multicomponential discretization models, named Models A–E, where each model identifies and presents a positive/negative diagnosis based on the experiences of the latest financial statements from six different industries. The varied components of the model test performance measurements comparatively by using data-preprocessing, data-discretization, feature-selection, two data split methods, machine learning, rule-based decision tree knowledge, time-lag effects, different times of running experiments, and two different class types. The experimental dataset had 24 condition features and a decision feature EPS that was used to classify the data into two and three classes for comparison. Empirically, the analytical results of this study showed that three main determinants were identified: total asset growth rate, operating income per share, and times interest earned. The core components of the following techniques are as follows: data-discretization and feature-selection, with some noted classifiers that had significantly better accuracy. Total solution results demonstrated the following key points: (1) The highest accuracy, 92.46%, occurred in Model C from the use of decision tree learning with a percentage-split method for two classes in one run; (2) the highest accuracy mean, 91.44%, occurred in Models D and E from the use of naïve Bayes learning for cross-validation and percentage-split methods for each class for 10 runs; (3) the highest average accuracy mean, 87.53%, occurred in Models D and E with a cross-validation method for each class; (4) the highest accuracy, 92.46%, occurred in Model C from the use of decision tree learning-C4.5 with the percentage-split method and no time-lag for each class. This study concludes that its contribution is regarded as managerial implication and technical direction for practical finance in which a multicomponential discretization model has limited use and is rarely seen as applied by scientific industry data due to various restrictions.