Toggle navigation
ReVival
Display
Datasets
Recovery
Synthetic data
Matrix data
Real-world data
Online recovery
RecovDB
Trend analysis
Prediction
Trend extraction
Centroid Decomposition
Decomposition calculator
Streaming decomposition
Sign vector strategies
About
Iteration: - | Frobenius distance: -
Recovery
using synthetic data
Recovery progress
show details
hide details
Showing recovery progress of matrix X as a chart (left) and in values (right)
Use the controls on the top to start, pause and reset the recovery progress
All missing values are in the base series
Values and chart of X are updated after each iteration of the recovery algorithm
The recovery terminates when Frobenius distance drops below the threshold
Unhide iteration progress for more details about the each iteration of the recovery algorithm
Frobenius distance
show details
hide details
Showing calculated Frobenius distance between X
_{i-1}
and X
_{i}
after each iteration i
The recovery progress runs until the Frobenius distance drops below the threshold epsilon = 0.01
Sign Vector Computation
show
hide
show details
hide details
Showing computation of the maximizing sign vectors for the current iteration
Notice how the maximizing sign vectors of the previous iteration are used as initialization
Z
_{1}
Z
_{2}
Z
_{3}
Z
_{4}
Z
^{T}
*V
show details
hide details
Showing the product Z
^{T}
*V of each sign vector as it maximizes through the SSV
Log
eXascale Infolab
March 2019