Global Data Science Forum

Expand all | Collapse all

Machine Learning

  • 1.  Machine Learning

    Posted Mon September 30, 2019 08:46 AM

    I have a csv dataset and i want to use  two or more clustering algorithms, build an unsupervised time-series classifier to identify characteristic day-length patterns. in csv data. in csv dataset each of the columns in the csv data set includes sensor measurements of the same kind for light in a room (units in Lux). I want to use appropriate quantitative metrics to determine the number of time series clusters and to evaluate their quality. In light of the data and the differences between algorithms, i want to speculate on why a given method yielded quantitatively better clusters. I have used many different Prerequisites but I've not been successful to do so.. some where in directional i am going wrong .. any idea ?
    in my dataset :↓

    ' Data columns (total 15 columns):
                       300000 non-null int64
        Date           300000 non-null object
      d1               300000 non-null int64
     d2                300000 non-null int64
      d3               300000 non-null int64
      d4               300000 non-null int64
      d5               300000 non-null int64
      d6               300000 non-null int64
      d7               300000 non-null int64
      d8               300000 non-null int64
      d9               300000 non-null int64
     d10               300000 non-null int64
     d11               300000 non-null int64
     d12               300000 non-null int64
     d13               300000 non-null int64
    dtypes: int64(14), object(1)
    memory usage: 34.3+ MB
    ~~~
    ' # sum1forline
    Elapsed time :  0.07872253399997135
    n =  300001
    
    
    # lenopenreadlines
    Elapsed time :  0.08887843000004068
    n =  300001
    
    
    # lenpd
    Elapsed time :  0.35768378400010986
    n =  300001
    
    
    # csvreaderfor
    Elapsed time :  0.4113426979999349
    n =  300001
    
    
    # openenum
    Elapsed time :  0.08758717699995032
    n =  300001
    ~~~
    
    ' df.shape 
    
    (300000, 15)
    ~~~
    
    'data.shape 
    
    (300000, 13)
    ~~~


    ------------------------------
    Reza Hashemi
    ------------------------------


  • 2.  RE: Machine Learning

    Posted Thu October 17, 2019 01:35 PM
    Hi Reza, have you learned of the specialized services in IBM Cloud ? I am referring to Watson Studio, whch is an IDE for Data Science.
    Here is the link:
    https://www.ibm.com/uk-en/cloud/watson-studio
    You can freely create an IBM Cloud account with Lite plan meaning no costs.
    https://cloud.ibm.com/login

    Models you can build using various instruments: code oriented (Python and R in Jupyter Notebooks, where you can use the scikit learn library), but also the very new, graphical and ML automatized model building tool, developed by IBM Research: Auto AI. This, besides the better known SPSS Modeler, which is available using Watson Machine Learning service in conjunction with Watson Studio.
    AutoAI uses various algorithms for the task selected, at the end of the processing, several models being transparently being compared.
    The Data being used to train the model can be uploaded to a Cloud Object Storage to which the mentioned services connect.
    This, aside from the possibility to feed real time data, with message hubs and streaming analytics, which you can also provision.
    Let me/us if you need more information.

    ------------------------------
    Viktor Kaznovsky
    ------------------------------



  • 3.  RE: Machine Learning

    Posted Fri October 18, 2019 08:12 AM

    Hi guys 👋🏿

    Princewill here,
    I'm  new here.
    I'll be interested in group projects in machine learning. 

    Thanks



    ------------------------------
    Princewill Okechukwu
    ------------------------------



  • 4.  RE: Machine Learning

    Posted Thu October 17, 2019 01:35 PM
    Hi Reza, have you learned of the specialized services in IBM Cloud ? I am referring to Watson Studio, whch is an IDE for Data Science.
    Here is the link:
    https://www.ibm.com/uk-en/cloud/watson-studio
    You can freely create an IBM Cloud account with Lite plan meaning no costs.
    https://cloud.ibm.com/login

    Models you can build using various instruments: code oriented (Python and R in Jupyter Notebooks, where you can use the scikit learn library), but also the very new, graphical and ML automatized model building tool, developed by IBM Research: Auto AI. This, besides the better known SPSS Modeler, which is available using Watson Machine Learning service in conjunction with Watson Studio.
    AutoAI uses various algorithms for the task selected, at the end of the processing, several models being transparently being compared.
    The Data being used to train the model can be uploaded to a Cloud Object Storage to which the mentioned services connect.
    This, aside from the possibility to feed real time data, with message hubs and streaming analytics, which you can also provision.
    Let me/us if you need more information.

    ------------------------------
    Viktor Kaznovsky
    ------------------------------



  • 5.  RE: Machine Learning

    Posted Mon November 11, 2019 09:47 PM
    Very interesting, but I see no mistake.
    But I am still learning the basics.

    ------------------------------
    PLPinto.M.
    ------------------------------