Home » Uncategorized

Remarkable Applications of Data Science

We are living in a time of slow yet steady insurgence of data science and AI in our lives. It spans more industries than we’d expect.

Over the past decade, Data Science has stretched out into almost every industry. Form industries like Automobiles and Healthcare to Finance as well as the Gaming Sector. It plays a significant part in the government sectors.  So here are some instances of the most impactful applications of data science.

Project Soli

One of the most significant technological transitions was the switch from buttons to touch screen. It transformed the operation of phones and computers. However, we’re moving on to the next big thing, touch-free phones.

Approved by the federal communications commission (FCC), Google has started working upon “Project Soli”. Project Soli is a system that is designed to track hand movements from a millimetre away by using miniature motion sensor radar. This system will enable people to control devices from TV’s to smartphones through gesture control.

How does it work?

Soli’s sensor technology emits electromagnetic waves in a broad beam. If an object falls within the radius of the beam, it scatters the energy of the electromagnetic waves. This causes a certain amount of the energy to reflect at the radar antenna. Elements such as energy, time delay and frequency shift within the reflected signal capture important information about the characteristics and dynamics of the object.

The radar technology uses high-frequency radio waves to detect moving objects which, according to Google, is profoundly more accurate than gesture tracking cameras. The sensor chips in Soli are capable of capturing up to 10,000 frames per second.

Predictive Policing

This application of data science is going to seem like its straight out of the popular Sci-Fi franchise “Robocop”. Law enforcement has been leveraging data science to enhance policing by using a system called predictive policing. Predictive policing is an AI-based system that harnesses big data, geospatial technologies and evidence-based intervention models to come up with algorithms that will increase public safety and reduce crimes.

How does it work?

Predictive policing works just like any analytical tool used in business industries to predict future market conditions and industry trends over time. This system analyses variables like demographic trends, people, place, groups, incidents, economic conditions, crime rates of particular areas. Based on the analysis of these variables, the system comes up with future predictions of criminal behavioural patterns. This empowers law enforcement to be proactive and respond more quickly to situations.

CrimeScan and Predpol are two predictive policing tools that have been developed for the U.S. Police Forces. However, the software can only forecast the possible crime vicinity. They do not identify the likely criminal suspects. In India, the Delhi Police joined forces with the ISRO (Indian Space Research Organization) to develop Crime Mapping Analytics and Predictive System (CMAPS). Through this system, the Delhi Police hope to have more stringent control over law and order by analyzing criminal behavioural patterns.

AI Fashion Designer

The juggernaut of e-commerce Amazon has been working on digitalizing its fashion designing methods. Amazon has developed a machine learning system that can analyze images of clothes and capture its style. It then creates an algorithm that is capable of designing new clothing from scratch based on its analysis.

How does it work?

The program was developed by their research centre based in San Francisco called Lab126. A tool called generative adversarial network (GAN). GAN consist of two deep neural networks that work as an overall system. The system analysis the properties of particular clothing’s style or design by scanning a series of photographs of it. GAN applies that design onto an existing piece of clothing, creating a completely new part of fashion from scratch.

Amazon revealed this program in one of its workshops is still in its developmental phase. The application of data science, in this case, might pave the way towards a future of automated fashion designing.

Digital Targeted Advertisements 

This is one of the most commonly used yet quite a smart application of data science. Every single one of us has come across these types of ads on social networking sites, online shopping sites, etc. These ads are not just random pop-up ads but are very specifically targeted.

How does it work?

Whenever an individual browses through the internet, the user’s browser generates browsed information to third-party advertising networks. The data is then stored in a small piece of code called browser cookies. This information is available to ad networks. The ad networks analyze the users’ behavioural pattern and accordingly target the individual based on their browsing habits.

Facebook Image Recognition

Face recognition is one of the most popular features of AI. Facebook has utilized data science to develop an algorithm of face recognition for its users. This feature will recognize a facial image in a photo even if the person has not been tagged. It will then notify the spotted individual of their appearance in the picture.

How does it work?

The first and most crucial step in this process is that the user needs to turn on the option of “tag suggestions” or “face recognition” under the photo privacy settings. Once this is done, Facebook’s machine learning system creates a template of the face in the image by analyzing its pixels. The template is created out of a string of number combinations.

No two templates are the same as the number coding sequence is different for every face. One the template is created, Facebook analysis the faces in the uploaded image and suggests a tag if it’s a match. When it comes to privacy concerns, the setting can be turned on and off as per the user’s requirement. Once the configuration is turned off, the created is automatically deleted.

Price Comparison Websites

This is another everyday use of data science these days. Price Comparison Websites are a standard tool that people use in the case of purchasing insurance or booking hotels as well. This stands as another example of data science simplifying things for our use.

How does it work?

Comparison sites obtain data for comparison through two methods, which are Crawling and Data Feeds.

Crawling: This method uses software which is programmed to visit and analyze merchant websites. The software then captures the prices of the products. The data is then used for comparison.

Data Feeds: This is a more straightforward way most sites use. In this method, the Merchant sites themselves prepare data files about their product’s prices and submit it to the Comparing sites.

The machine learning system then chronologically arranges the prices from the least to the most for its viewers.

Speech Recognition Assistants

Speech Recognition Assistants is the next wave of technology on the horizon. These digital voice assistants are designed to make our lives extremely simple and easy. With a voice assistant, one can command a computer to perform tasks. From tasks like casual internet browsing to quickly responding to emails, these assistants have got us all covered up.

How does it work?

When we speak, our voice generates small sound packets called “phones”. These “phones” are the sounds of letters or collective letters, which are words. As we speak to the voice assistant, the machine learning system analyses each “phone” in our words. It then digitalizes the words into a format the software can read. Once in a readable format, it analysis the data for its meaning. Based on algorithms, the AI makes an accurate educated guess as to what the speaker is saying. It then programs the smart device it is linked with to perform the desired task.

Airline Route Planning

Airline companies have leveraged the use of data science for several crucial tasks. One such task involves the planning of flight routes.

How does it work?

When it comes to planning a flight route, several factors come into play. Elements such as weather conditions, air traffic, popular flight routes, and runway bandwidth are majorly taken into consideration while planning an aircraft’s course. The machine learning systems analyze all this big data that is collected. The AI develops an algorithm based on these macro-economic factors. This algorithm is then used to create an entire flight route for the aircraft.

Moonshot thinking is vital with innovations. It allows people to go that stretch their imagination and work backwards. We live in a time where Moonshot thinking is invested in and encourages failures. This thinking leads to better discoveries. As the field expands, so will the applications of Data Science.

Is this something which you find yourself interested in ?

If you’d like to work at the forefront of innovative technology and use data to improve systems, I’d recommend you check out these jobs with one of the top technology consulting companies in Singapore and the Asia-Pacific region:

Sources: