Machine Learning 2.b : Applications

By Sagar Gandhi on

In the last post, we saw few applications of Machine Learning. continuing on the same grounds, more applications are presented in this post.

Just as a reminder, the point is to get us motivated enough so that we actually see how to do it ourselves.

(3)Marketing: In the age of globalization, bringing buyers and sellers together for mutually adventurous exchange is not an easy job. Simply because of the diversity of the influence factors, such as economic trends, social forces, political conditions, etc. Machine Learning is already so deeply involved that people cannot even imagine not using it for the Marketing.

  1. Market Research: Task Done: Market Research is the most essential aspect of the marketing. It not only gives the overview of customer’s wants, needs and beliefs, it also helps discovering how they act. Determining marketing strategy, understanding prices of commodities, analyzing supply demand situation, understanding the market trends and adapting to the requirements are only few basic applications of Market Research. Machine Learning helps in grounding this multidimensional space under single hood.
    Who: Various companies product wise, as well as service wise, for e.g. TechValidate, InterQ
  2. Advertising:
    Though Advertising is not the only component involved in the Marketing, it is a big and imperative piece of a pie. The advertising industry is gaining ground due to advantages that Machine Learning provides. Below are few of the major species:

i. Personalized Advertising:
Task Done: Machine Learning is heavily used to analyze the customers, perform predictive analytics and personalized messaging of a product customer actually needs. This also involves Market Segmentation and prediction of Customer Lifetime Value (CLV). The carried out Market Survey immensely helps in deciding actionlets.
Who: Rocket Fuel uses a methodology called “Moment Scoring” to predict what actions can be taken, with a particular person, at a particular time.

ii. Bidding
Task Done: In an Real-Time Bidding (RTB) environment, Demand-Side Platform (DSP) must figure out amount to be spent on a specific impression, in less than 100 milliseconds. The various performance metrics such as click-through rate (CLR) or conversion rate (CR) are assessed optimally using Machine Learning algorithms.
Who: Obviously, Google is one of the most telling players when it comes to RTB.

Besides these, other aspects of Marketing namely Product Pricing, Sales Strategy, etc. are gaining the involvement of Machine Learning.

(4) Smart City Solutions: Major technological, economical and environmental changes have stemmed the interest in smart cities. Smart cities contain the potential of fundamentally changing our lives with respect to many aspects such as less pollution, faster transportation, easy parking spot detection and of course, smart interactions. Machine Learning is at the heart of technologies used.

  1. Computer Vision:
    Task Done: Extracting information from the images is one of the most challenging tasks. Machine Learning, with the help of Image Processing techniques helps achieving this feat. The fundamental tasks like Object Detection, Biometric Identification and Tracking have enormous applications’ scope, be it security, interactions or smart management of the spatial and temporal information. Who: Sighthound: The company provides security camera software for home and businesses. They are also capable of detecting, tracking, and recognizing people and objects.

  2. Internet of Things (IoT):
    Task Done: The popular buzzword of the year, it continues to produces data full of treasures. The task, however, boils down to analyzing this data in real time, be it giving real-time insights to doctors or be it an alarm triggering maintenance of a device. What would be better tool than Machine Learning to realize this inhuman task?
    Who: There are multiple companies, varying with respect to the application domain. It would be unfair to list only a few.

Besides the listed domains, there are many other areas where Machine Learning is used heavily. Listing down all of them will drift us from the main purpose of learning How to do Machine Learning?. Just for the sake of completeness, below are the other areas where Machine Learning is getting applied.

Robotic Locomotion, Interaction centers at banks, airports and public places, Handwriting Recognition, Document Digitization, Song Prediction, Smart Training Rooms, Passenger Prediction, Music Generation, Crime Detection, Interview Rating, Attrition Prediction, Art History, Gait Detection and Tracking, Music Taste Prediction, Speech Recognition, Product Launch Success Rate Prediction, Cheminformatics, Brain-Machine Interfaces, etc ..

In addition to the mentioned ones, if you think some interesting application area is missed, please mention it in the comments. This way others will be able to find the current application areas in one place.

As it must be quite obvious by now, we are surrounded by Machine Learning, from listening songs to predicting next generation medicines. As a techno-enthusiast, it is not only inspiring but also essential to understand the tid-bit of Machine Learning.

From the next post onwards, we will delve into how things are actually done!

Note: I am thankful to Vivek, he took a time out of a busy schedule to review this post.