Objective: Inpatient length of stay (LOS) is an important measure of hospital activity, health care resource consumption, and patient acuity. This research work aims at developing an incremental expectation maximizati...
详细信息
Objective: Inpatient length of stay (LOS) is an important measure of hospital activity, health care resource consumption, and patient acuity. This research work aims at developing an incremental expectation maximization (EM) based learning approach on mixture of experts (ME) system for on-line prediction of LOS. The use of a batchmode learning process in most existing artificial neural networks to predict LOS is unrealistic, as the data become available over time and their pattern change dynamically. In contrast, an on-line process is capable of providing an output whenever a new datum becomes available. This on-the-spot information is therefore more useful and practical for making decisions, especially when one deals with a tremendous amount of data. Methods and material: The proposed approach is illustrated using a real example of gastroenteritis LOS data. The data set was extracted from a retrospective cohort study on all infants born in 1995-1997 and their subsequent admissions for gastroenteritis. The total number of admissions in this data set was n = 692. Linked hospitalization records of the cohort were retrieved retrospectively to derive the outcome measure, patient demographics, and associated co-morbidities information. A comparative study of the incremental learning and the batch-mode learningalgorithms is considered. The performances of the learningalgorithms are compared based on the mean absolute difference (MAD) between the predictions and the actual LOS, and the proportion of predictions with MAD < 1 day (Prop(MAD < 1)). The significance of the comparison is assessed through a regression analysis. Results: The incremental learningalgorithm provides better on-line prediction of LOS when the system has gained sufficient training from more examples (MAD = 1.77 days and Prop(MAD < 1) = 54.3%), compared to that using the batch-mode learning. The regression analysis indicates a significant decrease of MAD (p-value = 0.063) and a significant (p-value =
Clustered linear regression (CLR) is a new machine learning algorithm that improves the accuracy of classical linear regression by partitioning training space into subspaces. CLR makes some assumptions about the domai...
详细信息
Clustered linear regression (CLR) is a new machine learning algorithm that improves the accuracy of classical linear regression by partitioning training space into subspaces. CLR makes some assumptions about the domain and the data set. Firstly, target value is assumed to be a function of feature values. Second assumption is that there axe some linear approximations for this function in each subspace. Finally, there are enough training instances to determine subspaces and their linear approximations successfully. Tests indicate that if these approximations hold, CLR outperforms all other well-known machine-learningalgorithms. Partitioning may continue until linear approximation fits all the instances in the training set-that generally occurs when the number of instances in the subspace is less than or equal to the number of features plus one. In other case, each new subspace will have a better fitting linear approximation. However, this will cause over fitting and gives less accurate results for the test instances. The stopping situation can be determined as no significant decrease or an increase in relative error. CLR uses a small portion of the training instances to determine the number of subspaces. The necessity of high number of training instances makes this algorithm suitable for data mining applications. (C) 2002 Elsevier Science B.V. All rights reserved.
Multimodal fusion for identity verification has already shown great improvement compared to unimodal algorithms. In this paper, we propose to integrate confidence measures during the fusion process. We present a compa...
详细信息
An intrusion detection system (IDS) aims to increase the security of a computer system by dynamically monitoring various features and parameters of the system so as to be able to detect intrusions at the earliest poss...
详细信息
An intrusion detection system (IDS) aims to increase the security of a computer system by dynamically monitoring various features and parameters of the system so as to be able to detect intrusions at the earliest possible. IDS's have been developed for privileged UNIX programs like sendmail, lpr, and login. The IDS that we have built is for applets. It serves as a protection against malicious applets and warns the user when such applets are downloaded. Our system monitors applets using system call traces from the Java runtime environment. Feature vectors created from the system call traces are used to train a machine learning algorithm. The rule-set produced by the algorithm can then be used to distinguish hostile applets from good applets. (C) 2001 Elsevier Science Inc. All rights reserved.
Knowledge management is crucial to the teaching and learning process in the current era of digitalization. The idea of "learning via working together" is making Natural Language Processing a popular tool to ...
详细信息
Knowledge management is crucial to the teaching and learning process in the current era of digitalization. The idea of "learning via working together" is making Natural Language Processing a popular tool to improve the learning process based on the intelligent system for evaluating the composition. English language learning is highly dependent on the composition written by the students under various topics. Teachers are facing huge difficulties in the evaluation of the composition as the level of writing by the students will vary for individual. In this research, Natural Language Processing concept is utilized for getting trained with the student's writing skills and Multiprocessor learningalgorithm (MLA) combined with Convolutional Neural Network (CNN) (MLA-CNN) for evaluating the composition and declaring the scores for the students. The model's composition scoring rate is validated using a range of learning rate settings. Some theoretical notions for smart teaching are proposed, and it is hoped that this automatic composition scoring model would be used to grade student writing in English classes. When applied to the automatic scoring of students' English composition in schools, the suggested composition scoring system trained by the MLP-CNN has great performance and lays the groundwork for the educational applications of ML inside AI. The study results proved that the proposed model has provided an accuracy of 98%.
暂无评论