04
- January
2019
Posted By : Alexander Goida
Data Science Group level up

I’m very exciting to tell you that with Data Science Society of Sofia we’re launching a continuous event for everyone who wants get real experience with Data Science. You definitely need to got to the first meeting to know more. Subscribe here to not miss it. You will learn more about format and tasks which you can choose to work on. But even if you don’t have time to work on tasks this is still may suite you. Come and check it out by your self. A little more about it further in this article.

This activity doesn’t expect that you know Python, Jupyter Notebook, or statistics. You’ll get familiar with it during practicing some tasks. If you’re programmer who wants to move to this area, if you are a tech lead and need some insights, or even if you’re CTO and passioned about technical details this activity for you.

You can participate in two ways:

  • As a listener. This is read-only mode for the most busy people who just interesting in following up the progress of the group.
  • As a participant. This is the mode when you’re trying to follow along predefined milestones.

You will be given a list of tasks which you can choose to work on. There are different structure of the activity depends on the task. If you decide to chose mine, we’ll follow a sequence of development phases which are pretty common for Data Science projects. If you don’t like suggested tasks you may choose your own one and still follow the group getting all assistance as usual.

Tasks that I will lead

First of all, this is not a learning course with a teacher. So you won’t get a home work or predefined strong class structure. Also nobody will control your activity and pace. You’re the teacher and the boss for your self.

Together with you we’ll build a recommendation system using food reviews and learn how to develop on Python, prepare data for model training, create a model, build ETL pipeline and process to refresh existing models. But if you think it’s too complex for you, I prepared other tasks which you can choose. Below you can find description for each of them.

LevelProblemsTitleDescription
BeginnerSupervised Learning, Regression
Big Mart Sales
The data scientists at BigMart have collected 2013 sales data for 1559 products across 10 stores in different cities. Also, certain attributes of each product and store have been defined. The aim is to build a predictive model and find out the sales of each product at a particular store.
IntermediateClassification, ClusteringWaist-mounted smartphone trackerThe aim is to build a model by which a smartphone can detect its owner’s activity precisely. The data is built from the recordings of 30 subjects performing activities of daily living (ADL) while carrying a waist-mounted smartphone with embedded inertial sensors.
AdvancedSentiment analysis, classification, clustering, recommenderAmazon Fine Food RecommendationThe aim is to build a model for a recommendation system based on data from Amazon Fine Food Reviews.

If you like everything above and still didn’t decided about your participation, you can still join the Data Science Group on Facebook to stay tune.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.