Data Science – The Sexiest Job of the 21st Century

- RAJDEEP PATHAK

UG1

Though Data Science is what’s ruling the globe today, there is no formal definition to it. According to Wikipedia, “Data science is an inter-disciplinary field that uses scientific methods, processes, algorithms and systems to extract knowledge and insights from many structural and unstructured data.”

Data Science is the study of data – it is collecting and analyzing real world data to gain valuable insights that is of great use for finding a new approach for an organization. Data Science involves the use of Big Data, Machine Learning algorithms, and even Deep Learning models.

Big Data

Here’s the formal definition of Big data by Ernst and Young. “Big data refers to the dynamic, large and disparate volumes of data being created by people, tools and machines; it requires new, innovative and scalable technology to collect, host and analytically process the vast amount of data gathered in order to derive real-time business insights that relate to consumers, risk, profit, performance, productivity management and enhanced shareholder value.”

Artificial Intelligence (AI) vs Data Science (DS)

• Data Science is the process and method for analyzing large sets of data and extracting valuable insights out of them.

• Data Science involves mathematics, statistical analyses, data visualization, machine learning and more.

• Data Science tasks involve the use of Machine Learning algorithms as well as Deep Learning models. For example, a Regression model can be used to predict the price of a car, given its features in the dataset.

• Data Science is a broader term that takes in the complete data processing methodology.

• AI includes everything that allows computers to learn how to solve problems and make intelligent decisions on their own.

• AI can be thought of as an umbrella, covering Machine Learning and Deep Learning. Machine Learning is a subset of AI.

Data Science Methodology

Following steps is a common outline for data science methodology.

• Business Understanding: No doubt, the first step is business understanding. Data Science methodology begins with spending the time to seek clarification, to attain Business understanding.

• Analytic Approach: This step involves seeking clarification about the Business understanding, to find out the most appropriate path for analysis. The appropriate analytic approach for the problem is opted for in the context of the business requirements.

• Data requirements: The Data Scientist needs to identify what is the right data needed for solving the problem. How to collect them from the right source is also the issue addressed in this step. It is really vital to define the data requirements.

• Data Collection: After the data ingredients are collected, the Data scientist has a good understanding on the data he will be working with. Descriptive statistics and visualization techniques are applied to the data set to access initial insights about the data.

• Data Understanding: The stage of understanding data comprises all activities related to constructing the dataset.

• Data Preparation: This is one of the most crucial steps in the entire Data Science Methodology. This is the stage where data is transformed. It is the process of getting the data into a state where it becomes easier to work with. Missing entries and noises in the data set are handled in this step.

• Modelling: Then comes the most important step, which is modelling. Data modelling focuses on developing models based on the dataset. The models are either descriptive or predictive. The models are basically based on the analytic approach taken in step 2. It is either statistically driven, or Machine Learning driven.

• Model Evaluation: Evaluation is calculating the accuracy of the model. That is, how accurately it will work on the data on which it is not trained on. This is also known as out-of-sample accuracy.

• Model Deployment: Once the model is built and evaluated, it is deployed. Then it is put to ultimate test before it is brought to real world.

• Feedback: Once in play, feedback from the users helps refining the model and making it more accurate every time. The value of the model will depend on successfully incorporating the feedback.

This is an iterative step – Modelling, deployment, feedback. After the model is deployed, its performance is tracked and feedback is generated. This is important to keep the model improving, for a better out-of-sample accuracy.

Conclusion

There is no doubt why Data Science is regarded as the “sexiest job in the 21st century” by Harvard. It does not matter which field you come from, you can be a data scientist. The foremost attributes that one should have in order to be a good data scientist is that, he should be:

• Curious

• Judgmental

• Argumentative

• A good storyteller

The most important skills for Data Science include:

• Statistics and Probability

• Linear Algebra

• Python & R Programming

• Databases and SQL

The search for skilled data scientists by companies is on, which is increasing as days are passing by. No doubt, Data Science is revolutionizing (and will continue to do so) how companies work.