![]() ![]() For example, in one aspect, there is provided a processor implemented method for classifying multi-dimensional time series of parameters. SUMMARYĮmbodiments of the present disclosure present technological improvements as solutions to one or more of the above-mentioned technical problems recognized by the inventors in conventional systems. However, training RNNs requires large labeled training data like any other deep learning approach, and can be computationally inefficient because of sequential nature of computations. diagnoses, mortality prediction and estimating length of stay, and fault diagnostics from sensor data from machines and the like. With various parameters being recorded over a period of time in databases, Recurrent Neural Networks (RNNs) can be an effective way to model the sequential aspects of EHR data, e.g. There has been a growing interest in using deep learning models for various clinical prediction tasks from Electronic Health Records (HER), for medical diagnosis, to predict future diseases in patients, to predict unplanned readmission after discharge, and also for health monitoring of devices/machines, etc. The disclosure herein generally relates to time series analysis, and, more particularly, to systems and methods for classification of multi-dimensional time series of parameters. The entire contents of the aforementioned application are incorporated herein by reference. patent application claims priority under 35 U.S.C. Mapping from parameters to target class is considered while constraining the linear model to use only subset of large number of features. ![]() Extracted features are concatenated to learn a non-temporal linear classification model and weight is assigned to each extracted feature during learning which helps to determine relevant parameters for each class. A fixed-dimensional feature vector is outputted using a pre-trained unsupervised encoder, which acts as off-the shelf feature extractor. Embodiments of the present disclosure implement learning models for classification tasks in multi-dimensional time series by performing feature extraction from entity's parameters via unsupervised encoder and build a non-temporal linear classifier model. Building classification models requires large labeled data and is computationally expensive. Traditional systems and methods have implemented hand-crafted feature extraction from varying length time series that results in complexity and requires domain knowledge. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |