With data becoming the key, the position of data science has gained immense traction among companies. Businesses are heavily relying on analytics and are looking to hire data professionals who have relevant skill sets of managing complex business problems. In fact, despite this crisis, according to recent news, many companies are still hiring data scientists to work on innovative solutions that can help customers in this uncertain time.
The numbers have revealed that the requirement for data science and analytics practitioners in the US is projected to boom by 364,000 openings to 2,720,000, by 2020. Even for India, reports have stated the proportion of jobs (against global open jobs) advertised in India has risen from 7.9% in February 2020 to 9.2% in May.
Some of those prominent names include Airtel X Labs, PayPal, Oracle, Paytm, JP Morgan, among others. Majority of these companies usually opt for data scientists with some relevant experience in domain knowledge in order to get a heads up on their projects. However, many companies are also looking for junior or entry-level data scientists in order to hire them for lesser salaries and train them according to their company’s vision.
Sign up for your weekly dose of what's up in emerging technology.
But it is never easy to land on to these jobs, and that’s why it is critical for junior professionals to know what to expect during a data science interview in these companies and would require a specific kind of preparation. In this article, we will share a few things that will guide you to understand what sort of questions can be asked in an interview for an entry-level data scientist.
Apart from general questions about the candidate and their career graph, data science interviews involve a lot of technical questions. A majority of the job of a data scientist would include executing technical skills by analysing data and statistics. These questions are asked so that recruiters can get a perspective of the hands-on technical capabilities of candidates, and thus, would involve questions on mathematics, statistics, coding and machine learning. For this, the aspirants have to showcase training and learnings they have done over the years to grasp the knowledge of data science.
Answering these technical questions correctly would highlight the necessary knowledge and skills in statistics, databases and programming that candidates have and their relevance to the job requirement. These technical questions are also asked so that companies can understand the algorithmic knowledge of candidates as well as their expertise to use it. Many so-called data scientists usually push a lot of jargon on their resume, but during the interview, they fail to showcase their technical talent, and that’s why data scientists should never overstate their skills and should be prepared to answer some technical questions during the interview.
Some technical questions that could be asked during interviews are: –
- Popular libraries like computer vision, NLP and deep learning
- Various evaluation metrics for ML algorithms
- Python vs R
- Latest and popular ML, DL models and their workings
- ML-based cloud computing services of AWS, Azure and how are these used
- Mathematics behind every classification and regression algorithms
Project Specific Questions
Projects are a crucial part of a data scientist’s work, and therefore highlighting that segment in the interview would be key for junior data scientists to land on a job. And that’s why it is always suggested that data science aspirants must focus their efforts on working on projects based on computer vision, GANs, and more, in order to build their portfolio. Apart from working on projects, it is also critical for candidates to have a good knowledge of project details and its outcome because it would be expected by interviewers to ask technical questions based on the mentioned projects.
Project-specific questions would help interviewers to analyse the capabilities of candidates in handling complex business problems. And that’s why data scientists must prepare these pointers beforehand, and try to answer them in the clearest and most concise manner. These project-specific questions can include why candidates chose a particular project, the technical challenges while working on it, techniques used in cleaning the dataset, time taken for this project, the societal value of the project, tools used as well as the team involved. Highlighting details to interviewers would showcase the stronghold candidates have on his projects, and thus, would provide a competitive edge among others.
Understanding Of Soft Skills
Once interviewers get an understanding of candidates’ technical abilities and project knowledge, it is time for them to analyse the soft skills of candidates. Although soft skills are often not taken seriously by data scientists, in reality, soft skills are the skill sets that will help data scientists simplify their work and make it more understandable for the rest of the organisations. In fact, in many cases, soft skills turn out to be more important than expertise in technical skills. Google has done an internal survey across its teams and has noted that the best teams weren’t the technical specialists, instead, they were the ones that “brought strong, soft skills to the collaborative process.”
Having soft skills are critical for data to communicate with other teammates of organisations as well as sharing data insights to stakeholders. Some of the essential soft skills include communication, curiosity, business acumen, data storytelling, critical thinking and adaptability and product understanding. Alongside, enhancing soft skills will also help data scientists to understand the ethics of using data and its business value. And, analysing these factors would help interviewers to understand the vision, imagination and creativity of candidates.
Skill Of Data intuition
Another critical aspect that is usually analysed by interviewers and recruiters during the data science interview is the data intuition of candidates. Here the intuition is not the gut feeling of being creative with the data and the ability to mould the data according to the requirements. It is the instinctive understanding of the concepts and then applying the same while solving business problems. This skillset helps in analysing the massive amount of data in such a perspective that can be beneficial in solving business problems.
In fact, in an interview with Analytics India Magazine, Sanjiv Kumar Jha, Chief Data Scientist at Quantela, stated that data intuition is the most critical skill for data scientists. He said it is easy for data scientists to pick up technical expertise in their process, but the skill of data intuition cannot be acquired in a few days. “The crucial part of a data science workflow is the attitude of problem-solving from a data perspective,” said Jha.
The skill of data intuition would involve the understanding of when to be sceptical about a method and when to try different ways for problem-solving. Further, while working with a large amount of data, data intuition skills can help data scientists to understand and analyse it in a particular way. And therefore a junior data scientist or the professionals who are just starting out their career should remember that successful data scientists aren’t the ones who know all the maths, but it is somebody who knows when and how to use the maths and algorithms to gain business value.
Software Development Mindset
Lastly, while hiring data scientists, a software development mindset is one aspect where candidates are assessed on, where interviewers see how fierce these aspiring data scientists are while developing software and dealing with the challenges that come with coding and programming. Mindset is a collective skill, which includes the attitude of data scientists and the way they approach a business problem. Having a software developer’s mindset will help in making an informed decision while facing programming challenges and dealing with the failures.
Furthermore, software development mindset requires the understanding of patience and perseverance, which are critical to building complex software. With this skill, data scientists can have a better understanding of projects, which in turn will enable them to develop superior software solutions.