Hobby data science project ideas for a beginner?

projects

#1

Hi all,

I am Praveen. currently working in telecom domain.

I have started learning data analytics recently. I wanted to get some project ideas from all of you you. I want to do some hobby project which should have a market value.

What would you suggest me? I appreciate your valuable suggestions.


Project Ideas On Machine Learning
#2

you can try on Kaggle or try to analysis of Facebook trend or tweeter’s tweets analysis
All the best… :grinning:


Where can I find some free telecom data
#3

Hi Praveen,

Here are a few projects you can look at doing on the side:

If you are specifically interested in telecom data sets, here are a few of them:

  1. Here is a dataset / example on bigML platform for predicting customer churn. I believe this is one of the biggest challenge telecom players across the globe face today. So, this should definitely have a lot of market value. https://bigml.com/user/francisco/gallery/dataset/5163ad540c0b5e5b22000383

  2. There are a few datasets available for Milano and a few other cities. There are call logs data and other interaction data, which can be downloaded. https://dandelion.eu/datamine/open-big-data/

If you are complete beginner, I would recommend that you start with a few generic projects where you can get help from larger community. Once you have spent some time and got some expertise, you can look at the telecom datasets and projects.

Regards,
Kunal


#4

Hi Kunal, thanks for the reply. I wanted to do a hobby project which should have a market value. something like a app based on data analytics. kindly suggest few of the ideas which can be built as a app.


#5

Check these out:

For data in Indian context:

https://data.gov.in/


#6

Hi Kunal,

Thank you sharing great ideas and links.

Could you please share some project ideas from the links you mentioned (https://data.gov.in/) on what we can start working on. For e.g taking data sets of “travel and tourism” or “education”, what problem we can solve?

Abhishek Das