"Evaluating Anomaly detection methods"

cazzi123cazzi123 MemberPosts:3Contributor I
edited June 2019 inHelp
Hi,

I am trying to evaluate anomaly detection methods. I want to do this using ROC and F-measure. I am not sure how to do this. i have attached the xml of the process.

i would appreciate any assistance on this as I am still learning Rapidminer.

Thanks:) 















<运营商激活= =“false”类"anomalydetection:Local Correlation Integeral (LOCI)" compatibility="2.1.002" expanded="true" height="76" name="Local Correlation Integeral (LOCI)" width="90" x="179" y="390"/>

















<操作符= " true " class = " select_attribute激活s" compatibility="5.3.015" expanded="true" height="76" name="Select Attributes (2)" width="90" x="581" y="30">












<操作符= " true " class = " select_attribute激活s" compatibility="5.3.015" expanded="true" height="76" name="Select Attributes (3)" width="90" x="581" y="210">







<运营商激活= =“false”类"compare_rocs" compatibility="5.3.015" expanded="true" height="76" name="Compare ROCs" width="90" x="899" y="435">





<运营商激活= =“false”类"generate_attributes" compatibility="5.3.015" expanded="true" height="76" name="Generate Attributes (4)" width="90" x="313" y="390">




<运营商激活= =“false”类"set_role" compatibility="5.3.015" expanded="true" height="76" name="Set Role (3)" width="90" x="447" y="390">




<运营商激活= =“false”类"select_attributes" compatibility="5.3.015" expanded="true" height="76" name="Select Attributes (4)" width="90" x="581" y="390">




<运营商激活= =“false”类"anomalydetection:Generate ROC" compatibility="2.1.002" expanded="true" height="130" name="Generate ROC (3)" width="90" x="715" y="390">












<操作符= " true " class = " select_attribute激活s" compatibility="5.3.015" expanded="true" height="76" name="Select Attributes (5)" width="90" x="581" y="570">
















<连接from_op = from_po“生成属性(2)”rt="example set output" to_op="Set Role" to_port="example set input"/>




























[\code]
Tagged:

Answers

  • frasfras MemberPosts:93Contributor II
    Your value "3.5" for the threshold delivers no values "outlier" in the
    label attribute "Outlier_Label". I added a macro "threshold" to handle this more comfortable
    (lowering it to "1"):








    threshold
    1











    <运营商激活= =“false”类"anomalydetection:Local Correlation Integeral (LOCI)" compatibility="2.2.000" expanded="true" height="76" name="Local Correlation Integeral (LOCI)" width="90" x="179" y="390"/>

















    <操作符= " true " class = " select_attribute激活s" compatibility="6.0.003" expanded="true" height="76" name="Select Attributes (2)" width="90" x="581" y="30">












    <操作符= " true " class = " select_attribute激活s" compatibility="6.0.003" expanded="true" height="76" name="Select Attributes (3)" width="90" x="581" y="210">







    <运营商激活= =“false”类"compare_rocs" compatibility="6.0.003" expanded="true" height="76" name="Compare ROCs" width="90" x="899" y="435">





    <运营商激活= =“false”类"generate_attributes" compatibility="6.0.003" expanded="true" height="76" name="Generate Attributes (4)" width="90" x="313" y="390">




    <运营商激活= =“false”类"set_role" compatibility="6.0.003" expanded="true" height="76" name="Set Role (3)" width="90" x="447" y="390">




    <运营商激活= =“false”类"select_attributes" compatibility="6.0.003" expanded="true" height="76" name="Select Attributes (4)" width="90" x="581" y="390">




    <运营商激活= =“false”类"anomalydetection:Generate ROC" compatibility="2.2.000" expanded="true" height="130" name="Generate ROC (3)" width="90" x="715" y="390">












    <操作符= " true " class = " select_attribute激活s" compatibility="6.0.003" expanded="true" height="76" name="Select Attributes (5)" width="90" x="581" y="570">
















    <连接from_op = from_po“生成属性(2)”rt="example set output" to_op="Set Role" to_port="example set input"/>




























Sign InorRegisterto comment.