In addition, at WiDS Pune study group we have developed exhaustive content on Data Visualization and Exploratory Data Analysis.
1) Github and Kaggle are top 2 resources.
2) Networking within the organization and outside the organization through attending conference, meetups, having memberships of relevant associations, forums, bodies who are active in your domain.
3) Subscribing to social media groups, channels and regularly following them up.
4) Extensive reading through various mediums such as books, social media, news channels, forums, associations.
5) Participating via blogging, writing, speaking, competitions, hackathons.
6) W.r.t Data Science specifically, there are promotions at all levels – Companies like IBM, FB, MS, Google, Intel, Amazon etc have their own developer relation programs through which they give regular updates about their products, features, platforms, conduct workshops, even provide incentives and awards.
7) There are around 72 co-working spaces, Accelerators and Incubators in Pune who regularly organize events.
8) Following the experts, influencers in the field is also very important.
9) Harvard, Stanford release numerous free newsletters, podcasts, there are numerous you tube channels.
10) Taking up regular certifications.
11) Participating in various platforms like hackerranks, Makeover Mondays and other forums who throw various challenges.
12) Being aware of all aspects of the solution, product design such as law, privacy, policies, safety, Dev-Ops, ML-Ops, Productions, software engineering, solution design in addition to core technology.
13) Being able to add creativity, innovative excellent UX and CX to solution is also important.
1) Do you want to make a drastic and dramatic shift in your career? This requires a serious investment of time & money too, mostly a very formal certification would help or support from organization to directly start working in data science projects and learning on the job can also be an alternative. This is what we called re-skill! Dropping, what we are currently doing and taking up completely new skills and working on them for 8 to 10 hrs with a specific focus, relentlessly for 8 to 10 months is required.
2) If you want to taste the waters, and understand what you like or what you don’t like, whether you can do this, you may first start involving & attending meetups/online courses etc. Start understanding different aspects different sub domains like NLP, Image recognition, Object Recognition, conversational AI, so on. Each is a specialization in itself and needs in-depth ramp up. You may choose based on what you like, or find easy to pick up to gain confidence, or want to focus only on 1 sub skill like stats, maths, EDA, DV etc.
3) Since AI/ML/DS is a multi-disciplinary skill, some basics of Maths and Stats are a must and without understanding these we cannot jump into modelling. Lot of courses are available. Most important is to learn to apply.
4) If one is a good programmer, its very very beneficial as ramping up on Python can be very very beneficial. Some people feel more confident coding and generating output using Python libraries. Almost all code is available. 100s of cheat sheets, guidelines, checklists etc easily floating on internet.
5) Having tremendous patience for data, deep desire for research, constant study, read is a must. Being very comfortable to play with data, slice and dice the data in different ways, to derive meaning from data, make friends with data, treat the data is very important before deriving insights. Most important, it’s an iterative process. 70% of time goes in data engineering. Having good business knowledge, market knowledge, domain knowledge, subject matter expertise regarding the project is must. Only coding or calling libraries or apis will not yield outcome if we have no understanding of the data, patterns, relations, co-relations and so on. Start moving from IDEs to Jupiter notebooks.
6) Then many tools are available with AWS, Azure, Google, Intel and IBM Watson if you don’t want to code 1st and understand the ML pipeline. It totally depends individual comfort level and where to start and how to start.
7) There are many online trainings and some give free basic courses like - Coursera, Udemy, Analytics Vidhya, Medium, DataQuest and so on – Refer Learning resource on widspune.com
8) There are many platforms to practice such as hackerranks as it has test cases also, makeovermonday, programmiz etc.
9) There are always hackathon running almost on daily basis somewhere in all parts of the world, where we can participate and learn to solve a problem given the data sets.
10) In order to be a data scientist, we need to have a strong GitHub and Kaggle presence and profile. We need understand the difference between a Data Scientists and Data Engineer. Its same like computer science and computer engineering. To begin with, we may just start with data analytics or business analytics kind of role. Most Important is start somewhere and lot of self-study and self-motivation, determination and discipline is required and it’s an iterative process and no one can teach you everything and things are evolving as it’s a new domain.