7 New Year Resolutions That Data Science Aspirants Should Take Up

Every new year guarantees two things — fireworks and resolutions. Similar to fireworks, resolutions also extinguish after its momentary sparkles. Nevertheless, we try. For those who belong to the data science or machine learning community, these resolutions can be quite fun.

Here are a few resolutions for data science aspirants that are worth trying this new year:

Hype Is Real: 100 Days Of ML Code Challenge Works

No matter how much theory you are familiar with, data science mastery comes with hands-on practice. And, there are always forums with groups who are creating a 100 days ML code challenge almost every day. Data science aspirants can join one these to keep track of what other members of the group are up to, and also share your progress by displaying python notebooks or by sharing your GitHub links.

Maintain A Blog

The data science community never gets tired of reading beginner's guides. So, use this opportunity to share your learning via a blog that helps you gain credibility within the data science circles. There are several data science communities online in all social media platforms, and therefore, contributing to these communities will help in positioning oneself to be a part of the community and also be in constant touch with the peers of the community.

Don't Fall In Love With The Tools

There is always this big talk about whether one should opt for Python or R; Tableau or Power BI. All the industry practitioners, Kaggle grandmasters and experts recommend to be obsessed with the problem and not the tools. One needs to use the tool and play around with the problem by spending some sandbox time with it before solving the problem. Data science can help in overcoming these problems; however, many use fancy tools to solve a non-existed problem. Applying tools and frameworks depends on the problem statement, available data, implementation constraints and the expected solution. One needs to have a preference for solving problems rather than getting too worked up about languages, tools and frameworks.

Frame Your Own Problems

Charles Kettering once said, "A problem well defined is a problem half-solved." Framing the problems have always been the biggest challenge of data science, and like any other research process framing the problem is very important, i.e. stating one's hypothesis, and developing a strategic framework to solve the problems that are posed. Formulating better questions and defining a problem well help in solving the problem using data science approach and generate business insights and drive actionable plans.

Compete Wherever You Can

Competitions for data scientists are a great way to exercise and test one's machine-learning skills on real data and prediction tasks. One should always apply for competitions wherever they can, as opting for a competition by downloading the data and using the right tool and fitting complex framework can be a valuable learning experience for data science enthusiasts. Competitions will help enthusiasts to get their hand dirty on real data and provide a chance to deal with real-world machine learning problems in a structured environment. Competitions also offer substantial cash prizes, which in turn can help data science aspirants to improve their future. 

Keep Github Updated

GitHub is a site popular among data scientists that host code repositories, data, and interactive explorations; and having a profile on GitHub is very crucial to have a career in this field. Whether one is an aspirant or a pro data scientist, one needs to use Github for collaboration, making changes to projects, safely; and being able to track and rollback those changes over time. Updating GitHub is equally essential, as 99% of the companies that can hire you will use Git and GitHub, and an updated profile will make you more hireable and would set you apart from other developers.

Watch Out For Snake Oil Merchants

According to experts, snake-oil merchants have taken different form over the years and have entered the realms of AI and other advanced fields. One has to beware of these snake oil merchants by being aware of the implications of inappropriate usage of AI. AI for medical diagnosis or even detecting spam on email is fine because given enough data, a machine learning model can improve and be made reliable. However, in the case of ethical, social dilemmas such as public policing or criminal prediction, anomalies emerge and can prove fatal, and one has to be well known about it.