Being a python programmer, I have decided to get in the field of data science and later on AI and Machine Learning.
I have already brushed up the knowledge of statistics and other python libraries to be used in Data Science.
My question is, what's the best way to get into this field?
Should I start analysing the data directly on Kaggle, or there is a better approach of doing this?
Thanks!
Since you are a programmer, getting your hands dirty with Scikit-learn along with a cleaned Kaggle Dataset will get you far ahead. You also get to see how others also approach the problem. Once you get comfortable with the working pipeline, start exploring on your own with unlabelled datasets.
Dear Harshit, Kaggle can be a good approach for introducing into the matter, then you should study and analize with your own methods (e.g. fitting of data) and tools (e.g. a numerical software) the datasets of your interest. Gianluca
Dear Harshit,
Matlab is worldwide used for prototyping and experimenting. Moreover, it a valuable way to review the field. Follow and investigate from https://ch.mathworks.com/help/stats/index.html
Best Regards
Hi Stéphane,
Thanks for the deep insight.
I am already quite comfortable with programming in python, so is switching to Matlab a good idea?
Harshit.
Matlab is a vector and matrix language that requires some adaptation time on starting. Neverthless, you may just use online Matlab resources to familiarize with Big Data Analytics and Machine Learning concerns. Matlab is a reference.
Thanks for your reply.
Dear Harshit and Stephane, only for "equity", I would suggest you the power of numerical methods of Mathematica; for Data Science, see e.g. the following link. Gianluca
http://www.wolfram.com/solutions/industry/data-science/?source=nav
Better to continue your work with python as the scope for python looks amazingly better. You could start analyzing things through the usage of classifier for data classification
pip install biopython
Dear Harshit,
I think Kaggle seems great, but you should also check other things. For example, for a starter, you should check simple problems in "Plasma Membrane Protein Trafficking" or other problems like bioeconomics. Do not stick to one range of problems.
Also, check out the Biopython module. Version 1.7 has a big improvement.
Best regards,
Mohammadreza Niknam Hamidabad
Since you are a programmer, getting your hands dirty with Scikit-learn along with a cleaned Kaggle Dataset will get you far ahead. You also get to see how others also approach the problem. Once you get comfortable with the working pipeline, start exploring on your own with unlabelled datasets.
Inlining Felan Carlo Garcia's answer, Scikit-learn provides a very useful documentation about Machine Learning in Python. Follow http://scikit-learn.org/stable/_downloads/scikit-learn-docs.pdf
In addition, I recommend you to consult the numerous and valuable "NumPy / SciPy Recipes" for Data Sciences and Machine Learning by Christian Bauckhage on ResearchGate.
Regards
Your mentioned Data Science fields is a young one. I suggest WEKA tools and R Programming language is a better one for handling data science fields.
Reference Book.
1. Data Science and Big Data Analytics published by EMC Educational Service,Wiley
2. Introductory R - Robert J. Knell
- Badges
- Science topic