A Novel HOSFS Algorithm for Online Streaming Feature Selection

Sandhiya, S.; Palani, U.

doi:10.1109/icscan49426.2020.9262401

Cited by 3 publications

(2 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In second step subset selection is performed by PSO Algorithm it is used to evaluate all the selected subsets and selects the best subset from the generated n number of subset. Sandhiya S and Palani U [14] proposed an online feature selection algorithm for streaming features called Heuristic Online Streaming Feature Selection (HOSFS) which has advantages on choosing features from streaming features and omits the irrelevant and redundant features in real-time by using self-adaption sliding window protocol, and Heuristic function. Which assigns heuristic value to the features using the trained heuristic function and selects features with higher heuristic value where other features are considered as irrelevant features.…”

Section: Related Workmentioning

confidence: 99%

An Efficient Disease Prediction System with Electronic Health Record Using Feature Selection and Convolutional Recurrent Neural Network

Sandhiya

2023

Preprint

View full text Add to dashboard Cite

Rapid advancements in communication technologies and expert systems generate massive amounts of medical data. Using an electronic health record (EHR) and a disease prediction system, we can predict high-risk patients at an early stage. The use of EHR for early prediction aids in the improvement of health care quality through personalized medicine. To analyze the massive amount of disease-related EHR data, various deep learning methods for disease prediction are proposed. Deep learning is a machine learning technique that has evolved. Even with more deep learning techniques in place, no healthcare system has achieved a higher level of disease prediction accuracy. In this paper, we propose a disease prediction system for serious, lethal diseases such as heart, diabetes, and cancer diseases. The proposed system is made up of a feature selection algorithm (GBCOA) and a Convolutional-Recurrent Neural Network (C-RNN). This proposed model was tested against existing methods and found to have higher prediction accuracy with less computation time.

show abstract

Section: Related Workmentioning

confidence: 99%

An Efficient Disease Prediction System with Electronic Health Record Using Feature Selection and Convolutional Recurrent Neural Network

Sandhiya

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The main difference between PRL and the just mentioned approaches is that, thanks to the feature generator component, PRL has the capability of treating problems with infinitely many features. PRL, like HOSFS [42], is also theoretically suitable for dealing with streaming of features, however, its main limitation is the efficiency that it could not be ideal in settings with high throughput.…”

Section: On-line Feature Selectionmentioning

confidence: 99%