Reporting and Analytics toolbox

Vilius,

I've been working with several tools for Data Science / Machine Learning.

• Knime, which has a whole bunch of different ML, DL algorithm and ETL tools available.   This is well respected in Gartner reviews and you can get a fairly competent version for free.  However, I find a few of its User Interface and UX decisions to be challenging.
• Weka which is a great place to learn ML approaches. It's basically a GUI front-end to a largish set of JAVA ML algorithms.  There is an interesting set of class from Future learn called Data Mining with Weka (and two follow-on courses) that are very useful.  The class can be audited for free.  This tool is also open source.  However, the tool is a little bit daunting without the class.
• Dataiku's DSS is my primary environment for a community membership project I'm working on with Science Museum of Minnesota (SMM) and *82 a data analytics group in the St. Paul Minnesota area.  This tool provides a GUI front end that uses "Mojito Cocktail Shaker" Scripts.  Shaker Scripts are a rather obscure ETL toolset.  But rather nice from my point of view. DSS uses scikit-learn python library for ML.  It also makes available R and Python programming.  All of which can be used with a modern web-based GUI, or in Jupyter Notebooks.  This toolset can be a pain to setup on PC.  Macintosh is a bit easier.  This tool can work with Big Data Solutions like Hadoop, Spark.  However, right now I'm using a PostgreSQL backend that seems to be working OK.  And yes, it's free as long as you don't need a direct MS SQL server connection. (They are willing to give non-profit organizations licenses for free or reduced pricing.)  My next step is to try to set up a copy on paperspace in the cloud in order to get more computer resources.  On the current project, with several hundred thousand examples of memberships, my 8GB core duo laptop is grinding to a halt trying to do the needed calculations.  In some cases taking several hours to evaluate only a 10% sample of the data.  :-(
• For ETL I really like Microsoft's Power Query inside of MS Excel. I have been able to successfully teach both marketing and fundraising staff to use this tool.  They got good enough at this that they have done presentations at TLCC on this subject.  I'm also using this for the current SMM project.  Microsoft makes a moderately powerful ELT and visualization tool available in Power BI Desktop.  This is not an ML or DL tool.  But a creditable BI tool until we get Sisense from the Tessitura Network as part of the Tessitura BI suite coming out this summer.  (Sisense is not expected to have any ML or DL components so the tools above or R will likely continue to have a place for the foreseeable future.)
• Finally, I just got access to Tableau for visualization.  This is a super powerful tool.  (very expensive) But wonderful.  It connects nicely to my PostgreSQL database making the visualization more productive than trying to do this in ggplot2.  Power BI, DSS, Knime all have visualization tools built in.  None are as good.  I do like R for correlation matrixes.  I've not figured out how to do a quick one line matrix in Tableau or any other tool.

Hope that this is a bit helpful to you.

Have a great day and weekend.  Let me know if you wish to talk at some time.  We could set up a Google Hangout or Zoom call.

Vilius,

I've been working with several tools for Data Science / Machine Learning.

• Knime, which has a whole bunch of different ML, DL algorithm and ETL tools available.   This is well respected in Gartner reviews and you can get a fairly competent version for free.  However, I find a few of its User Interface and UX decisions to be challenging.
• Weka which is a great place to learn ML approaches. It's basically a GUI front-end to a largish set of JAVA ML algorithms.  There is an interesting set of class from Future learn called Data Mining with Weka (and two follow-on courses) that are very useful.  The class can be audited for free.  This tool is also open source.  However, the tool is a little bit daunting without the class.
• Dataiku's DSS is my primary environment for a community membership project I'm working on with Science Museum of Minnesota (SMM) and *82 a data analytics group in the St. Paul Minnesota area.  This tool provides a GUI front end that uses "Mojito Cocktail Shaker" Scripts.  Shaker Scripts are a rather obscure ETL toolset.  But rather nice from my point of view. DSS uses scikit-learn python library for ML.  It also makes available R and Python programming.  All of which can be used with a modern web-based GUI, or in Jupyter Notebooks.  This toolset can be a pain to setup on PC.  Macintosh is a bit easier.  This tool can work with Big Data Solutions like Hadoop, Spark.  However, right now I'm using a PostgreSQL backend that seems to be working OK.  And yes, it's free as long as you don't need a direct MS SQL server connection. (They are willing to give non-profit organizations licenses for free or reduced pricing.)  My next step is to try to set up a copy on paperspace in the cloud in order to get more computer resources.  On the current project, with several hundred thousand examples of memberships, my 8GB core duo laptop is grinding to a halt trying to do the needed calculations.  In some cases taking several hours to evaluate only a 10% sample of the data.  :-(
• For ETL I really like Microsoft's Power Query inside of MS Excel. I have been able to successfully teach both marketing and fundraising staff to use this tool.  They got good enough at this that they have done presentations at TLCC on this subject.  I'm also using this for the current SMM project.  Microsoft makes a moderately powerful ELT and visualization tool available in Power BI Desktop.  This is not an ML or DL tool.  But a creditable BI tool until we get Sisense from the Tessitura Network as part of the Tessitura BI suite coming out this summer.  (Sisense is not expected to have any ML or DL components so the tools above or R will likely continue to have a place for the foreseeable future.)
• Finally, I just got access to Tableau for visualization.  This is a super powerful tool.  (very expensive) But wonderful.  It connects nicely to my PostgreSQL database making the visualization more productive than trying to do this in ggplot2.  Power BI, DSS, Knime all have visualization tools built in.  None are as good.  I do like R for correlation matrixes.  I've not figured out how to do a quick one line matrix in Tableau or any other tool.

Hope that this is a bit helpful to you.

Have a great day and weekend.  Let me know if you wish to talk at some time.  We could set up a Google Hangout or Zoom call.