Recommendations System with PySpark

[ujjwalll]

• Abstract (2-3 lines) Recommendations systems are everywhere around us, and with the explosion of data in the last few years, they are becoming more and more important. Every e-commerce website, whether it is youtube, Amazon, Netflix, google facebook, etc. They all using different recommendation algorithm to show us the best results out of all the enormous data set. This talk is mainly focused on how the recommendations systems are made, what are its applications, what are the different approaches, which approach is best. I will also cover my method which I have used in my project to make recommendations using machine learning in python.

• Brief Description and Contents to be covered I will mainly focus on how the basic recommendation system works. What are some machine learning techniques used in these systems and some latest techniques? For more information please refer slide 2 in the link provided below.

• Pre-requisites for the talk Basic knowledge of python and some idea of machine learning.

• Time required for the talk 30 mins.

• Link to slides link to slides

• Will you be doing hands-on demo as well? Yes, I will show my own project demo.

• Link to ipython notebook (if any) NA

• About yourself I am Ujjwal Singh, a first year, CSE - undergrad at IIIT Delhi. I am involved in machine learning and python from about 6-7 months, but I have learned a lot in these days. I have worked with Vidyamandir classes for their website development. I am also a open source contributor at FOSSASIA and PSF. I have completed projects like sentiment analyser for reviews for e-commerce websites in machine learning.

• Are you comfortable if the talk is recorded and uploaded to PyData Delhi's YouTube channel ? Yes

• Any query ? No .

[shagunsodhani]

Thank you @ujjwalll for proposing a talk. The outline looks good. I think the talk would be even more interesting if you could reduce the amount of text in the slides and increase the font. You could also consider adding some hands-on example so that the talk becomes more engaging.

[ujjwalll]

I have updated the slides according to suggestions. Please have a look at it Link for updated slides

Sorry for being late.

[shagunsodhani]

Looks pretty clean now. Best of luck.

[MSanKeys963]

@ujjwalll you want to deliver this in our upcoming meetup?

[ujjwalll]

Yes, sure.