Level Crossing Prediction with Neural Networks View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2009-08-13

AUTHORS

Halfdan Grage, Jan Holst, Georg Lindgren, Mietek Saklak

ABSTRACT

A level crossing predictor or alarm system with prediction horizon k is said to be optimal if it, at time t detects that an upcrossing will occur at time t + k, with a certain high probability and simultaneously gives a minimum number of false alarms. For a Gaussian stationary process, the optimal level crossing predictor can be explicitly specified in terms of the predicted value of the process itself and of its derivative. To the authors knowledge this simple optimal solution has not been used to any substantial degree. In this paper it is shown how a neural network can be trained to approximate an optimal alarm system arbitrarily well. As in other methods of parametrization, the choice of model structure, as well as an appropriate representation of data, are crucial for a good result. Comparative studies are presented for two Gaussian ARMA-processes, for which the optimal predictor can be derived theoretically. These studies confirm that a properly trained neural network can indeed approximate an optimal alarm system quite well – with due attention paid to the problems of model structure and representation of data. The technique is also tested on a strongly non-Gaussian Duffing process with satisfactory results. More... »

PAGES

623-645

References to SciGraph publications

  • 2005-04-08. An improved back propagation algorithm topredict episodes of poor air quality in SOFT COMPUTING
  • 1993. Statistical aspects of neural networks in NETWORKS AND CHAOS — STATISTICAL AND PROBABILISTIC ASPECTS
  • Identifiers

    URI

    http://scigraph.springernature.com/pub.10.1007/s11009-009-9153-3

    DOI

    http://dx.doi.org/10.1007/s11009-009-9153-3

    DIMENSIONS

    https://app.dimensions.ai/details/publication/pub.1035931193


    Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
    Incoming Citations Browse incoming citations for this publication using opencitations.net

    JSON-LD is the canonical representation for SciGraph data.

    TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

    [
      {
        "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
        "about": [
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/01", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Mathematical Sciences", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0104", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Statistics", 
            "type": "DefinedTerm"
          }
        ], 
        "author": [
          {
            "affiliation": {
              "alternateName": "Novo Nordisk A/S, Bagsv\u00e6rd, Denmark", 
              "id": "http://www.grid.ac/institutes/grid.425956.9", 
              "name": [
                "Novo Nordisk A/S, Bagsv\u00e6rd, Denmark"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Grage", 
            "givenName": "Halfdan", 
            "id": "sg:person.01167677057.51", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01167677057.51"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Mathematical Statistics, Lund University, Lund, Sweden", 
              "id": "http://www.grid.ac/institutes/grid.4514.4", 
              "name": [
                "Mathematical Statistics, Lund University, Lund, Sweden"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Holst", 
            "givenName": "Jan", 
            "id": "sg:person.012640460143.11", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012640460143.11"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Mathematical Statistics, Lund University, Lund, Sweden", 
              "id": "http://www.grid.ac/institutes/grid.4514.4", 
              "name": [
                "Mathematical Statistics, Lund University, Lund, Sweden"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Lindgren", 
            "givenName": "Georg", 
            "id": "sg:person.012274436413.00", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012274436413.00"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Visma Software AB, Limhamn, Sweden", 
              "id": "http://www.grid.ac/institutes/None", 
              "name": [
                "Visma Software AB, Limhamn, Sweden"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Saklak", 
            "givenName": "Mietek", 
            "type": "Person"
          }
        ], 
        "citation": [
          {
            "id": "sg:pub.10.1007/978-1-4899-3099-6_2", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1089745563", 
              "https://doi.org/10.1007/978-1-4899-3099-6_2"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/s00500-004-0435-y", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1000170180", 
              "https://doi.org/10.1007/s00500-004-0435-y"
            ], 
            "type": "CreativeWork"
          }
        ], 
        "datePublished": "2009-08-13", 
        "datePublishedReg": "2009-08-13", 
        "description": "A level crossing predictor or alarm system with prediction horizon k is said to be optimal if it, at time t detects that an upcrossing will occur at time t\u2009+\u2009k, with a certain high probability and simultaneously gives a minimum number of false alarms. For a Gaussian stationary process, the optimal level crossing predictor can be explicitly specified in terms of the predicted value of the process itself and of its derivative. To the authors knowledge this simple optimal solution has not been used to any substantial degree. In this paper it is shown how a neural network can be trained to approximate an optimal alarm system arbitrarily well. As in other methods of parametrization, the choice of model structure, as well as an appropriate representation of data, are crucial for a good result. Comparative studies are presented for two Gaussian ARMA-processes, for which the optimal predictor can be derived theoretically. These studies confirm that a properly trained neural network can indeed approximate an optimal alarm system quite well \u2013 with due attention paid to the problems of model structure and representation of data. The technique is also tested on a strongly non-Gaussian Duffing process with satisfactory results.", 
        "genre": "article", 
        "id": "sg:pub.10.1007/s11009-009-9153-3", 
        "inLanguage": "en", 
        "isAccessibleForFree": true, 
        "isPartOf": [
          {
            "id": "sg:journal.1043739", 
            "issn": [
              "1387-5841", 
              "1573-7713"
            ], 
            "name": "Methodology and Computing in Applied Probability", 
            "publisher": "Springer Nature", 
            "type": "Periodical"
          }, 
          {
            "issueNumber": "4", 
            "type": "PublicationIssue"
          }, 
          {
            "type": "PublicationVolume", 
            "volumeNumber": "12"
          }
        ], 
        "keywords": [
          "optimal alarm system", 
          "level crossing predictor", 
          "model structure", 
          "simple optimal solution", 
          "certain high probability", 
          "method of parametrization", 
          "time t", 
          "neural network", 
          "Gaussian stationary process", 
          "optimal solution", 
          "stationary processes", 
          "representation of data", 
          "optimal predictor", 
          "appropriate representation", 
          "false alarms", 
          "minimum number", 
          "network", 
          "representation", 
          "high probability", 
          "parametrization", 
          "system", 
          "problem", 
          "upcrossings", 
          "probability", 
          "solution", 
          "good results", 
          "satisfactory results", 
          "alarm system", 
          "terms", 
          "derivatives", 
          "alarms", 
          "results", 
          "prediction", 
          "technique", 
          "process", 
          "structure", 
          "choice", 
          "number", 
          "data", 
          "due attention", 
          "comparative study", 
          "values", 
          "degree", 
          "authors", 
          "attention", 
          "substantial degree", 
          "predictors", 
          "study", 
          "method", 
          "paper", 
          "crossing predictor", 
          "prediction horizon k", 
          "horizon k", 
          "optimal level crossing predictor", 
          "Level Crossing Prediction", 
          "Crossing Prediction"
        ], 
        "name": "Level Crossing Prediction with Neural Networks", 
        "pagination": "623-645", 
        "productId": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "pub.1035931193"
            ]
          }, 
          {
            "name": "doi", 
            "type": "PropertyValue", 
            "value": [
              "10.1007/s11009-009-9153-3"
            ]
          }
        ], 
        "sameAs": [
          "https://doi.org/10.1007/s11009-009-9153-3", 
          "https://app.dimensions.ai/details/publication/pub.1035931193"
        ], 
        "sdDataset": "articles", 
        "sdDatePublished": "2022-01-01T18:20", 
        "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
        "sdPublisher": {
          "name": "Springer Nature - SN SciGraph project", 
          "type": "Organization"
        }, 
        "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/article/article_491.jsonl", 
        "type": "ScholarlyArticle", 
        "url": "https://doi.org/10.1007/s11009-009-9153-3"
      }
    ]
     

    Download the RDF metadata as:  json-ld nt turtle xml License info

    HOW TO GET THIS DATA PROGRAMMATICALLY:

    JSON-LD is a popular format for linked data which is fully compatible with JSON.

    curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/s11009-009-9153-3'

    N-Triples is a line-based linked data format ideal for batch operations.

    curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/s11009-009-9153-3'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s11009-009-9153-3'

    RDF/XML is a standard XML format for linked data.

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s11009-009-9153-3'


     

    This table displays all metadata directly associated to this object as RDF triples.

    148 TRIPLES      22 PREDICATES      83 URIs      73 LITERALS      6 BLANK NODES

    Subject Predicate Object
    1 sg:pub.10.1007/s11009-009-9153-3 schema:about anzsrc-for:01
    2 anzsrc-for:0104
    3 schema:author Nac517d9b71b44f74801085e383d278ee
    4 schema:citation sg:pub.10.1007/978-1-4899-3099-6_2
    5 sg:pub.10.1007/s00500-004-0435-y
    6 schema:datePublished 2009-08-13
    7 schema:datePublishedReg 2009-08-13
    8 schema:description A level crossing predictor or alarm system with prediction horizon k is said to be optimal if it, at time t detects that an upcrossing will occur at time t + k, with a certain high probability and simultaneously gives a minimum number of false alarms. For a Gaussian stationary process, the optimal level crossing predictor can be explicitly specified in terms of the predicted value of the process itself and of its derivative. To the authors knowledge this simple optimal solution has not been used to any substantial degree. In this paper it is shown how a neural network can be trained to approximate an optimal alarm system arbitrarily well. As in other methods of parametrization, the choice of model structure, as well as an appropriate representation of data, are crucial for a good result. Comparative studies are presented for two Gaussian ARMA-processes, for which the optimal predictor can be derived theoretically. These studies confirm that a properly trained neural network can indeed approximate an optimal alarm system quite well – with due attention paid to the problems of model structure and representation of data. The technique is also tested on a strongly non-Gaussian Duffing process with satisfactory results.
    9 schema:genre article
    10 schema:inLanguage en
    11 schema:isAccessibleForFree true
    12 schema:isPartOf N2fa88c268a2748e086361b9f6becea19
    13 Ndaac8d176d90440ca657cf020e15d1c6
    14 sg:journal.1043739
    15 schema:keywords Crossing Prediction
    16 Gaussian stationary process
    17 Level Crossing Prediction
    18 alarm system
    19 alarms
    20 appropriate representation
    21 attention
    22 authors
    23 certain high probability
    24 choice
    25 comparative study
    26 crossing predictor
    27 data
    28 degree
    29 derivatives
    30 due attention
    31 false alarms
    32 good results
    33 high probability
    34 horizon k
    35 level crossing predictor
    36 method
    37 method of parametrization
    38 minimum number
    39 model structure
    40 network
    41 neural network
    42 number
    43 optimal alarm system
    44 optimal level crossing predictor
    45 optimal predictor
    46 optimal solution
    47 paper
    48 parametrization
    49 prediction
    50 prediction horizon k
    51 predictors
    52 probability
    53 problem
    54 process
    55 representation
    56 representation of data
    57 results
    58 satisfactory results
    59 simple optimal solution
    60 solution
    61 stationary processes
    62 structure
    63 study
    64 substantial degree
    65 system
    66 technique
    67 terms
    68 time t
    69 upcrossings
    70 values
    71 schema:name Level Crossing Prediction with Neural Networks
    72 schema:pagination 623-645
    73 schema:productId N9f208fb6f2e549cebf7d5a1a68d11604
    74 Ne85146dc90d1436d92deb0790ed25569
    75 schema:sameAs https://app.dimensions.ai/details/publication/pub.1035931193
    76 https://doi.org/10.1007/s11009-009-9153-3
    77 schema:sdDatePublished 2022-01-01T18:20
    78 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    79 schema:sdPublisher N3b82452b8c664ccaa1b6cdea4851dac2
    80 schema:url https://doi.org/10.1007/s11009-009-9153-3
    81 sgo:license sg:explorer/license/
    82 sgo:sdDataset articles
    83 rdf:type schema:ScholarlyArticle
    84 N2fa88c268a2748e086361b9f6becea19 schema:issueNumber 4
    85 rdf:type schema:PublicationIssue
    86 N3b82452b8c664ccaa1b6cdea4851dac2 schema:name Springer Nature - SN SciGraph project
    87 rdf:type schema:Organization
    88 N49b9298addff49458fcfc4a9989b57cc schema:affiliation grid-institutes:None
    89 schema:familyName Saklak
    90 schema:givenName Mietek
    91 rdf:type schema:Person
    92 N7e10a45ff1bd4ede8ac055ccff169532 rdf:first N49b9298addff49458fcfc4a9989b57cc
    93 rdf:rest rdf:nil
    94 N97ce5369419c4f0e936f3e73eabad4bf rdf:first sg:person.012640460143.11
    95 rdf:rest Nc2418bcd5f354b518d98c3fac6fe1fd7
    96 N9f208fb6f2e549cebf7d5a1a68d11604 schema:name doi
    97 schema:value 10.1007/s11009-009-9153-3
    98 rdf:type schema:PropertyValue
    99 Nac517d9b71b44f74801085e383d278ee rdf:first sg:person.01167677057.51
    100 rdf:rest N97ce5369419c4f0e936f3e73eabad4bf
    101 Nc2418bcd5f354b518d98c3fac6fe1fd7 rdf:first sg:person.012274436413.00
    102 rdf:rest N7e10a45ff1bd4ede8ac055ccff169532
    103 Ndaac8d176d90440ca657cf020e15d1c6 schema:volumeNumber 12
    104 rdf:type schema:PublicationVolume
    105 Ne85146dc90d1436d92deb0790ed25569 schema:name dimensions_id
    106 schema:value pub.1035931193
    107 rdf:type schema:PropertyValue
    108 anzsrc-for:01 schema:inDefinedTermSet anzsrc-for:
    109 schema:name Mathematical Sciences
    110 rdf:type schema:DefinedTerm
    111 anzsrc-for:0104 schema:inDefinedTermSet anzsrc-for:
    112 schema:name Statistics
    113 rdf:type schema:DefinedTerm
    114 sg:journal.1043739 schema:issn 1387-5841
    115 1573-7713
    116 schema:name Methodology and Computing in Applied Probability
    117 schema:publisher Springer Nature
    118 rdf:type schema:Periodical
    119 sg:person.01167677057.51 schema:affiliation grid-institutes:grid.425956.9
    120 schema:familyName Grage
    121 schema:givenName Halfdan
    122 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01167677057.51
    123 rdf:type schema:Person
    124 sg:person.012274436413.00 schema:affiliation grid-institutes:grid.4514.4
    125 schema:familyName Lindgren
    126 schema:givenName Georg
    127 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012274436413.00
    128 rdf:type schema:Person
    129 sg:person.012640460143.11 schema:affiliation grid-institutes:grid.4514.4
    130 schema:familyName Holst
    131 schema:givenName Jan
    132 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012640460143.11
    133 rdf:type schema:Person
    134 sg:pub.10.1007/978-1-4899-3099-6_2 schema:sameAs https://app.dimensions.ai/details/publication/pub.1089745563
    135 https://doi.org/10.1007/978-1-4899-3099-6_2
    136 rdf:type schema:CreativeWork
    137 sg:pub.10.1007/s00500-004-0435-y schema:sameAs https://app.dimensions.ai/details/publication/pub.1000170180
    138 https://doi.org/10.1007/s00500-004-0435-y
    139 rdf:type schema:CreativeWork
    140 grid-institutes:None schema:alternateName Visma Software AB, Limhamn, Sweden
    141 schema:name Visma Software AB, Limhamn, Sweden
    142 rdf:type schema:Organization
    143 grid-institutes:grid.425956.9 schema:alternateName Novo Nordisk A/S, Bagsværd, Denmark
    144 schema:name Novo Nordisk A/S, Bagsværd, Denmark
    145 rdf:type schema:Organization
    146 grid-institutes:grid.4514.4 schema:alternateName Mathematical Statistics, Lund University, Lund, Sweden
    147 schema:name Mathematical Statistics, Lund University, Lund, Sweden
    148 rdf:type schema:Organization
     




    Preview window. Press ESC to close (or click here)


    ...