Learning Specifications for Labelled Patterns

被引:2
作者
Basset, Nicolas [1 ]
Dang, Thao [1 ]
Mambakam, Akshay [1 ]
Jarabo, Jose Ignacio Requeno [2 ]
机构
[1] Univ Grenoble Alpes, VERIMAG, CNRS, Grenoble, France
[2] Western Norway Univ Appl Sci HVL, Dept Comp Math & Phys, Bergen, Norway
来源
FORMAL MODELING AND ANALYSIS OF TIMED SYSTEMS, FORMATS 2020 | 2020年 / 12288卷
关键词
Signal pattern matching; Monotonic specification learning; Pareto multi-criteria optimization; Signal Temporal Logic;
D O I
10.1007/978-3-030-57628-8_5
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
In this work, we introduce a supervised learning framework for inferring temporal logic specifications from labelled patterns in signals, so that the formulae can then be used to correctly detect the same patterns in unlabelled samples. The input patterns that are fed to the training process are labelled by a Boolean signal that captures their occurrences. To express the patterns with quantitative features, we use parametric specifications that are increasing, which we call Increasing Parametric Pattern Predictor (IPPP). This means that augmenting the value of the parameters makes the predicted pattern true on a larger set. A particular class of parametric specification formalisms that we use is Parametric Signal Temporal Logic (PSTL). One of the main contributions of this paper is the definition of a new measure, called c-count, to assess the quality of the learned formula. This measure enables us to compare two Boolean signals and, hence, quantifies how much the labelling signal induced by the formula differs from the true labelling signal (e.g. given by an expert). Therefore, the c-count can measure the number of mismatches (either false positives or false negatives) up to some error tolerance E. Our supervised learning framework can be expressed by a multicriteria optimization problem with two objective functions: the minimization of false positives and false negatives given by the parametric formula on a signal. We provide an algorithm to solve this multi-criteria optimization problem. Our approach is demonstrated on two case studies involving characterization and classification of labeled ECG (electrocardiogram) data.
引用
收藏
页码:76 / 93
页数:18
相关论文
共 30 条
  • [1] Quantitative Regular Expressions for Arrhythmia Detection
    Abbas, Houssam
    Rodionova, Alena
    Mamouras, Konstantinos
    Bartocci, Ezio
    Smolka, Scott A.
    Grosu, Radu
    [J]. IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS, 2019, 16 (05) : 1586 - 1597
  • [2] A THEORY OF TIMED AUTOMATA
    ALUR, R
    DILL, DL
    [J]. THEORETICAL COMPUTER SCIENCE, 1994, 126 (02) : 183 - 235
  • [3] [Anonymous], IMPLEMENTATION PARET
  • [4] [Anonymous], IMPLEMENTATION STLEV
  • [5] [Anonymous], ECGFIVEDAYS DATA SET
  • [6] [Anonymous], 2016, P 5 INT WORKSHOP SOF
  • [7] Asarin E., 2012, Runtime Verification: Second International Conference, RV 2011, San Francisco, CA, USA, September 27-30, 2011, Revised Selected Papers 2, V7186, P147, DOI [10.1007/978-3-642-29860-812, 10.1007/978-3-642-29860-8_12, DOI 10.1007/978-3-642-29860-8_12]
  • [8] ParetoLib: A Python']Python Library for Parameter Synthesis
    Bakhirkin, Alexey
    Basset, Nicolas
    Maler, Oded
    Jarabo, Jose-Ignacio Requeno
    [J]. FORMAL MODELING AND ANALYSIS OF TIMED SYSTEMS (FORMATS 2019), 2019, 11750 : 114 - 120
  • [9] Specification and Efficient Monitoring Beyond STL
    Bakhirkin, Alexey
    Basset, Nicolas
    [J]. TOOLS AND ALGORITHMS FOR THE CONSTRUCTION AND ANALYSIS OF SYSTEMS, PT II, 2019, 11428 : 79 - 97
  • [10] Efficient Parametric Identification for STL
    Bakhirkin, Alexey
    Ferrere, Thomas
    Maler, Oded
    [J]. HSCC 2018: PROCEEDINGS OF THE 21ST INTERNATIONAL CONFERENCE ON HYBRID SYSTEMS: COMPUTATION AND CONTROL (PART OF CPS WEEK), 2018, : 177 - 186