Data science or Data Science involves scientific methods, processes and systems to extract knowledge or a better understanding of data in its different forms, whether structured or unstructured, which is a continuation of some fields of data analysis such as statistics, data mining, machine learning, and predictive analytics. Despite dating from the 60s and 70s, it is in recent years when this discipline is having the greatest impact in the business world. In this post I am going to tell you what are the main knowledge that a data scientist should have, as well as what are its applications solving real problems in different industries. Are you interested? Well let’s go for it.

Data science and technology have always gone hand in hand, as technology cannot exist without computational technology to support it. In fact, “ data ” is a concept that emerged in the 1940s, the stage of the first computers, with the meaning of “information that can be transmitted and stored in a computer”.

From then on, the evolution of computers and their capacity to store and process data have shaped the concept of data analysis first and Data Science later, increasing the scope of analysis and the reliability of predictions .

However, Data Science is not just data and computers. This is centered on a figure, that of the Data Scientist.

Main knowledge that a Data Scientist must have.

To fully understand what data science can do for business, you first need to know what skills and knowledge data scientists need to master.

Although there is no clear and concise definition, there is an approximation made in 2010 by Drew Conway in which the concept of Data Science is reached from the areas of knowledge that it is necessary to master, and these are:

#1. Hacking Skills.

It refers to skills acquired to handle different types of data in different formats and for which there is no single integration method in a Data Science project. They are skills to “seek life” managing data sources that are not always structured, in search of relationships, predictions or useful patterns in a certain sector or business area.

#two. Substantive Expertise.

It concerns the knowledge of the area of ​​activity or the business from which the data comes. The same data, in different business areas, is also handled differently. For example, car driving data can be used to design predictive maintenance strategies in a workshop or to offer personalized insurance based on the way you drive.

#3. Disciplines such as Machine Learning or traditional analytical methods.

They are subsets of knowledge areas, while Data Science encompasses them all. They are pieces of a huge puzzle like Data Manipulation and Analysis.

However, in recent times, a new element has come into play: Artificial Intelligence . And it is that, as the computing power of computational systems has been increasing, AI has been emerging in parallel. Without going any further, IBM’s AI Watson faced real people in 2011 in the contest ‘Jeopardy’ using DeepQA (deep questions and answers) technology. Dozens of different algorithms are involved in it to process natural language, classify, search for relationships or categorize the statistical accuracy of the response. The AI ​​won the contest.

The “appearance” of intelligence emerges from hardware capable of doing calculations fast enough to make it appear as if we were dealing with a human contestant. Watson used distributed computing through Hadoop and databases that had to be stored in RAM for fast response.

9 applications of Data Science or Data Science in business.

Technological advances have led to the handling of very large amounts of data in very short times. In the same way, this has facilitated the ability to integrate these methods into user interfaces, making it more accessible to people and, therefore, to companies.

This, let’s call it, democratization of data, has meant that its use is spreading to different industries and sectors, in which it is providing quick and effective solutions to everyday problems that are faced in those markets every day.

The nine most powerful applications could be these:

#1. Cybersecurity: identification of cyberthreats

The detection is made from the access data to the systems and network resources. Patterns are searched for and the alert is given when situations that do not respond to a predefined pattern are detected.

The data comes from activity logs, with overwhelming amounts of data collected in historical files. From them, activity patterns are extracted to use as a reference.

#two. Finance: Fraud Detection

A similar process is applied, for example, in the detection of fraud in credit card payments. Here, systems can cross-reference data from different sources, such as a customer’s regular activity, along with “normal” usage.

In this way, it is possible to identify fraudulent scenarios (duplicate/stolen cards or improper/duplicate charges), paralyzing or warning of irregular activity before the damage occurs.

#3. Insurance: calculation of premiums

The insurance sector is another that benefits from Data Science. By analyzing driving habits using sensors, an insurance company can calculate a customer’s accident risks and offer a personalized quote for him. You can even introduce variable concepts that depend on analyzing your routines at different times of the year.

#4. Medicine: detection of tumors and search for treatments

Fields such as image analysis in the identification of diseases are perfect candidates to apply Data Science. When the images are obtained in a CT, X-ray or ultrasound, the recognition systems begin to be better even than the human specialists themselves.

To achieve such a high success rate, it is necessary to choose and process tens of thousands of scans to statistically train image recognition systems based on Supervised Machine Learning.

The same applies to the discovery of new drugs or to offer personalized treatments.

#5. Industry: predictive maintenance or the health of machines

Predictive maintenance is a clear example of the application of Data Science in industry. The machines, logistics systems and other elements of an industrial plant integrate thousands of sensors that collect data on temperatures, hours of operation, speeds, distances, noise level, etc.

A lot of information is generated that has to be prepared, filtered, cleaned and introduced into Machine Learning or Deep Learning models to predict failures in advance. As a consequence, substantial savings are achieved in periodic revisions or in the purchase of spare parts. Not to mention preventing a production plant from shutting down by surprise.

#6. Marketing: classification of customers and audiences

Currently, Data Science is capable of using social networks in real time as sources. In this way, it is possible to predict the demand for a product to create it from offers segmented by social class, cultural preferences, purchasing power, gender, hobbies…

In marketing departments, this data helps prepare reports prior to campaigns, launches or promotions.

#7. Search engines: image recognition

Take Google Photos as an example. On this platform, the photos we upload are automatically analyzed and classified based on those elements that Google’s AI is capable of identifying, whether they are cars, planes, people, flowers, food, animals, landscapes or unique places, among others.

Data Science intervenes in the choice of data (images) to train Deep Learning models. To realize its importance, remember that when Google was asked to search for gorillas, it returned photos of people of color as a result. Google initially solved this by removing “gorilla” from the search.

#8. Automation: cars that drive themselves

It is one of the most ambitious territories of Data Science. Automating car parking is not the same as automating complete driving, so there is still a long way to go on this path.

#9. Energy: securing supply

In the energy sector, Data Science is applied to different areas, such as predictive maintenance of its facilities and infrastructures and distribution networks, or consumption forecasting, to schedule energy generation tasks.

It is also used to detect fraudulent use of the grid ─such as illegal hookups─, prevent supply drops or real-time pricing.

As you can see, data science is increasingly important and represents a competitive advantage for any business, compared to its competition, not only because it is more efficient, but because it offers real-time solutions to those who really matter, which are consumers.