no code implementations • 9 May 2021 • Matthew Middlehurst, James Large, Gavin Cawley, Anthony Bagnall
We demonstrate that the temporal dictionary ensemble (TDE) is more accurate than other dictionary based approaches.
1 code implementation • 15 Apr 2021 • Matthew Middlehurst, James Large, Michael Flynn, Jason Lines, Aaron Bostrom, Anthony Bagnall
Since it was first proposed in 2016, the algorithm has remained state of the art for accuracy on the UCR time series classification archive.
no code implementations • 20 Aug 2020 • Matthew Middlehurst, James Large, Anthony Bagnall
We propose combining TSF and catch22 to form a new classifier, the Canonical Interval Forest (CIF).
no code implementations • 25 Apr 2020 • Anthony Bagnall, Paul Southam, James Large, Richard Harvey
Given the massive volume of luggage that needs to be screened for this threat, the best way to automate the detection is to first filter whether a bag contains an electric device or not, and if it does, to identify the number of devices and their location.
no code implementations • 13 Apr 2020 • Anthony Bagnall, Michael Flynn, James Large, Jason Lines, Matthew Middlehurst
The Hierarchical Vote Collective of Transformation-based Ensembles (HIVE-COTE) is a heterogeneous meta ensemble for time series classification.
no code implementations • 27 Nov 2019 • Anthony Bagnall, James Large, Matthew Middlehurst
We call this type of approach to TSC dictionary based classification.
no code implementations • 1 Nov 2018 • James Large, Paul Southam, Anthony Bagnall
tl;dr: no, it cannot, at least not on average on the standard archive problems.
1 code implementation • 31 Oct 2018 • Anthony Bagnall, Hoang Anh Dau, Jason Lines, Michael Flynn, James Large, Aaron Bostrom, Paul Southam, Eamonn Keogh
In 2002, the UCR time series classification archive was first released with sixteen datasets.
no code implementations • 18 Sep 2018 • James Large, Anthony Bagnall, Simon Malinowski, Romain Tavenard
We find that whilst ensembling is a key component for both algorithms, the effect of the other components is mixed and more complex.
no code implementations • 25 Oct 2017 • James Large, Jason Lines, Anthony Bagnall
We show that the Heterogeneous Ensembles of Standard Classification Algorithms (HESCA), which ensembles based on error estimates formed on the train data, is significantly better (in terms of error, balanced error, negative log likelihood and area under the ROC curve) than its individual components, picking the component that is best on train data, and a support vector machine tuned over 1089 different parameter configurations.
no code implementations • 28 Mar 2017 • Anthony Bagnall, Aaron Bostrom, James Large, Jason Lines
We describe what results we expected from each class of algorithm and data representation, then observe whether these prior beliefs are supported by the experimental evidence.
no code implementations • 4 Feb 2016 • Anthony Bagnall, Aaron Bostrom, James Large, Jason Lines
These algorithms have been evaluated on subsets of the 47 data sets in the University of California, Riverside time series classification archive.