Machine Learning Anomaly Detection with Python and Power BI

แชร์
ฝัง
  • เผยแพร่เมื่อ 23 ส.ค. 2024
  • Use the Isolation Forest algorithm to create unsupervised machine learning to identify outliers in your data. Leverage the power of Sckit Learn and Power BI
    Anomaly detection can be determined in a number of ways such as percentiles, quartile and standard deviation. These are all statistical devices that allow the user to set the threshold range for the expected data.
    #outliers
    #anomaly detection
    #powerbi
    #DAX

ความคิดเห็น • 17

  • @absentdata
    @absentdata  3 ปีที่แล้ว +2

    You can find the PBIX file and data files on the GitHub file. Please subscribe and like the video:
    github.com/Gaelim/Isolation-Forest

  • @agustinsantellanfernandez5587
    @agustinsantellanfernandez5587 2 ปีที่แล้ว +3

    Your videos are amazing, i am learning a lot of things that I use at work with direct impact, Please keep doing videos we appreciate your time and effort.

  • @padraichayes3836
    @padraichayes3836 2 ปีที่แล้ว

    Your videos are brilliant, the perfect amount of detail 👌🏻 thank you!

    • @absentdata
      @absentdata  2 ปีที่แล้ว

      Glad you like them!

  • @magnuswestrom3450
    @magnuswestrom3450 4 ปีที่แล้ว +2

    Nice video! I'm not an expert, so take my input with a grain of salt, but if I'd change anything it'd be to add the index-column before duplicating the query. Just as a best practice.

    • @absentdata
      @absentdata  4 ปีที่แล้ว +1

      Thanks for the tips!

  • @henkhbit5748
    @henkhbit5748 3 ปีที่แล้ว +1

    Great explanation how to combine python machine learning with power bi!

  • @thailinhpham6695
    @thailinhpham6695 3 ปีที่แล้ว +1

    Hi thank you for your great instruction. Could you mind make a understandable video to instruct about the temperature time series with RNN (recurrent neuron network). Thank you in advanced!

  • @arlistranmer1296
    @arlistranmer1296 2 ปีที่แล้ว

    Great video thank you!

    • @absentdata
      @absentdata  2 ปีที่แล้ว

      Glad you liked it!

  • @vikviking3000
    @vikviking3000 3 ปีที่แล้ว

    I am confused by something: why does one graphs have indexes on the x axis and the other one dates? Wouldn't it make more sense to have dates both times? Also it seems to me that what this algorithm does in this case is say that any point above the 95th percentile and below the 5th percentile are anomalies. Does it work well if the data has trends and seasonalities?

  • @jimmyrico5364
    @jimmyrico5364 3 ปีที่แล้ว

    This is great, thanks for sharing!
    Just one question, in the Power BI documentation says that to run Python on any dataset it has to be set as public which is sensitive if the data you’re using is private, any advice to overcome this challenge please?
    Appreciate any guidance and keep up the great work!

    • @absentdata
      @absentdata  3 ปีที่แล้ว +1

      You can duplicate the dataset and make any areas anonymous then link it back to
      the sensitive data index

  • @leonardorazoislas1688
    @leonardorazoislas1688 3 ปีที่แล้ว

    on 9:30 , "the script run with the followinb python instalation ~\Anaconda3\envs\py35." , whats mean? o have to set uo env variables?

    • @absentdata
      @absentdata  3 ปีที่แล้ว

      Please watch this video for help.
      th-cam.com/video/fZqzdbKrg2I/w-d-xo.html

  • @123jamalq
    @123jamalq 3 ปีที่แล้ว

    Can u please share the datasets?

  • @onurerdogan2236
    @onurerdogan2236 3 ปีที่แล้ว

    can you share dashboard template with me