Bottom Page

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
 Practical Machine Learning
#1
I have been working on machine learning either part time or full time for the last year and a half.

But what i am missing the practicality of it all.

I want to see a problem preferably a classification one worked from soup to nuts.

I just have not found one like that.

I read allot of the medium articles, but there is always something missing. The basic
python code that starts with the data, cleans it and normalizes and then
uses it in machine learning project.

I find so little of that in the literature.

What book or paper or website addresses this problem?

I just need to see how its done. I can do the proofs and I follow the math arguments, but where and how to I apply them?

Any help appreciated. Thanks in advance.

Respectfully,

ErnestTBass



Quote
#2
(Jul-14-2020, 03:39 AM)ErnestTBass Wrote: I read allot of the medium articles, but there is always something missing. The basic
python code that starts with the data, cleans it and normalizes and then
uses it in machine learning project.

I find so little of that in the literature.

What book or paper or website addresses this problem?

I think it's about knowing the data you have and how you intend to use it. For example, if you have time series data, there might be instrumental trends that are present and need to be removed. For images, you likely want to pick out a region of interest. At least in my case, this sort of knowledge came from exploring the data and understanding where it came from, etc. If you can read them, scientific papers tend to explain what preprocessing steps have been performed, from which you should be able to take the data, apply the same steps and reproduce the results. These days, I can imagine people are sharing code in, say, Jupyter notebooks, as open data is a growing thing.

Quote:I can do the proofs and I follow the math arguments, but where and how to I apply them?

I tend to think proofs and things are not to be applied when you're using the techniques, but rather are there to give one some appreciation and understanding of where the technique came from and why it works the way it does.
Quote

Top Page

Possibly Related Threads...
Thread Author Replies Views Last Post
  New to Machine Learning Rpereir1 1 402 Sep-21-2019, 05:17 PM
Last Post: ThomasL
  Suggestions for a machine learning model for invoice processing multiple languages? animeshsarraf 1 525 May-26-2019, 04:43 AM
Last Post: heiner55
  Do Machine Learning Employers care if I get a minor in Math on top of MS engineering AndrewScience 2 1,835 Oct-25-2017, 01:24 PM
Last Post: sparkz_alot

Forum Jump:


Users browsing this thread: 1 Guest(s)