Storage scaling management model

Introduction: The implementation of data storage process requires timely scaling of the infrastructure to accommodate the data received for storage. Given the rapid accumulation of data, new models of storage capacity management are needed, which should take into account the hierarchical structure of the data storage, various requirements for file storage and restrictions on the storage media size. Purpose: To propose a model for timely scaling of the storage infrastructure based on predictive estimates of the moment when the data storage media is fully filled. Results: A model of storage capacity management is presented, based on the analysis of storage system state patterns. A pattern is a matrix each cell of which reflects the filling state of the storage medium at an appropriate level in the hierarchical structure of the storage system. A matrix cell is characterized by the real, limit, and maximum values of its carrier capacity. To solve the scaling problem for a data storage system means to predict the moments when the limit capacity and maximum capacity of the data carrier are reached. The difference between the predictive estimatesis the time which the administrator has to connect extra media. It is proposed to calculate the values of the predictive estimates programmatically, using machine learning methods. It is shown that when making a short-term prediction, machine learning methods have lower accuracy than ARIMA, an integrated model of autoregression and moving average. However, when making a long-term forecast, machine learning methods provide results commensurate with those from ARIMA. Practical relevance: The proposed model is necessary for timely allocation of storage capacity for incoming data. The implementation of this model at the storage input allows you to automate the process of connecting media, which helps prevent the loss of data entering the system.

Download Full-text

Learning Data Correction for Myoelectric Hand Based on “Survival of the Fittest”

Cyborg and Bionic Systems ◽

10.34133/2021/9875814 ◽

2021 ◽

Vol 2021 ◽

pp. 1-12

Author(s):

Yusuke Yamanoi ◽

Shunta Togo ◽

Yinlai Jiang ◽

Hiroshi Yokoi

Keyword(s):

Machine Learning ◽

Data Storage ◽

Learning Performance ◽

Huge Amount ◽

Learning Methods ◽

Machine Learning Methods ◽

Wearable Robots ◽

Processing Data ◽

Survival Of The Fittest ◽

Learning Data

In recent years, myoelectric hands have become multi-degree-of-freedom (DOF) devices, which are controlled via machine learning methods. However, currently, learning data for myoelectric hands are gathered manually and thus tend to be of low quality. Moreover, in the case of infants, gathering accurate learning data is nearly impossible because of the difficulty of communicating with them. Therefore, a method that automatically corrects errors in the learning data is necessary. Myoelectric hands are wearable robots and thus have volumetric and weight constraints that make it infeasible to store large amounts of data or apply complex processing methods. Compared with general machine learning methods such as image processing, those for myoelectric hands have limitations on the data storage, although the amount of data to be processed is quite large. If we can use this huge amount of processing data to correct the learning data without storing the processing data, the machine learning performance is expected to improve. We then propose a method for correcting the learning data through utilisation of the signals acquired during the use of the myoelectric hand. The proposed method is inspired by “survival of the fittest.” The effectiveness of the method was verified through offline analysis. The method reduced the amount of learning data and learning time by approximately a factor of 10 while maintaining classification rates. The classification rates improved for one participant but slightly deteriorated on average among all participants. To solve this problem, verifying the method via interactive learning will be necessary in the future.

Download Full-text