Where is Data Science Used?
Learn from History, Build Future Projects.
Data Science is not a new field. Its being used from several centuries in solving various data related problems spanning across several domains. As the technology has evolved fast, in the recent times, it has made the Data Science methodologies accessible to a wide range of people. This evolution has further increased the reach and scope of Data Science applications to the most granular levels of business.
In this article we’ll discuss a course of Data Science applications, in chronological order. Learn how the technology advancement has connected the Programmers and Scientists and revolutionized the analytics. Along with some pointers to the future projects.
Data Science in History: Applications before IT
The discoveries and innovations made by most of our ancient scientists were all data driven.
Name it in any field:
- Space Research,
- Medical,
- Geography etc
All findings were preceded by a lot of data analysis. The scientists used to generate a lot of data by conducting various scientific experiments. The data collected from different scientific experiments was then enriched and reorganized by applying various mathematical theorems. Then derived the scientific inferences from the transformed data.
These scientific and statistically significant inferences were critical for the scientists to conclude their findings. These conclusions can be broadly classified into two types: Either 1) they determine the reason or cause behind a fact or 2) they generalize and explain how a system works. Generalization helps in predicting the future behavior of the system. This classification aligns with the outcome of solving any data science problem — Determine or Generalize.
As explained in Steve Burtons lecture [link provided in the playlist]: Tycho Brahe collected enormous amount of spatial data conducting several experiments. Kepler determined the planetary motion, by analyzing the data. Newton generalized the laws of motion, by analyzing the data.
While discussing today’s Applications we need to understand two things — the standard definition of Data Science and an IT definition of Data Science. In job world and IT industry, Data Science has acquired a specialized definition (though inline with the standard definition).
Any problem that involves data collection, application of statistical & mathematical theories and whose outcome is to determine a cause or generalize a pattern is a Data Science problem. IT industry extends this definition to a specialized scenario, which has become the default definition in Job market.
Information technology, of the current age, has brought several advancements in the computation methodologies. Distributed Computing and Parallel Processing have made a great difference in bringing down the computing time of large volume data. The time required to calculate mathematical and statistical functions, on a large volume & continuously growing dataset, is brought down from days to seconds and almost to real time. And machines are made to learn continuously from the business transactions.
This being the major value addition, IT world perceives majority of Data Science problems as Volume Intensive or Computationally Complex scenarios. Though it supports in solving the standard Data Science problems.
Data Science applications as per regular definition:
Every time a group of students submits assignments to the class teacher, the teacher collects some data about the performance of the students. After a few cycles, the teacher understands before hand which student is going to submit the assignment first, who is going to do it more accurate, who comes up with more clarification requests etc.
Here the Teacher is a Data Scientist who performs Generalization of the data by drawing some statistical inferences from the observations made during the previous cycles. Our brain has this intrinsic ability of apply mathematical functions on the observations we make.
A Parent knows, which food is allergic to the toddler and which one keeps the child healthy. Here the Parent is a Data Scientist by determining the cause (food item) of a fact (Health status) from the data collected (Health condition monitoring).
Like wise several experiments and researches being conducted by the students at various universities, to determine or generalize some thing with statistical evidence, can be exemplified as Data Science Applications.
Data Science applications as per IT specialized definition:
The IT definition of Data Science covers mostly the enterprises.
Data Science deals with every part of an organization that needs to determine a cause or generalize a behavior, by analyzing a large volume or variety of data and producing statistical evidences.
Collect the data, Determine the cause or Generalize a behavior. Where ever, these three things are required in business, administration or personal life that’s the area Data Science can play a role. Even if it cannot directly determine the cause, it will empower those who has the knowledge of determining why something has occurred.
It translates discrete data into meaningful and actionable interpretation, parses signal from noise and detects Patterns.
Systems could be of any size, micro to large scale. A simple web traffic analysis to a large scale airlines traffic controller. Data Science explains how a system works and generalizes how certain parts of the system are going to work in future based on some factors.
The following play list provides pointers to some innovative applications/thoughts of current time — small, medium & large.
Website traffic analysis: Log files written by the web programmers provide information about the users and the time & duration they have spent at each page. Combining this information with the information about the static controls at each page, a data scientist can program a layer of intelligence that determines the relationship between the color of an HTML Control and the Duration spent by a user at that page.
Based on this intelligence provided by the data scientist, the web programmer can change the color of a control to increase the user engagement at a critical page.
Global Forest Watch :
In this project, the deforestation is monitored in near real time. Data collected from several satellite images will be analyzed to generate statistical metrics. The underlying machine learning algorithms detect the anomalies in near real time and determines the patterns of deforestation.
Data Science Ahead : Near Future Applications
As computers become more and more faster, storage becomes unlimited and network improves in the security and bandwidth aspects, the Distributed Computing shall now be resulting in computations with more Precision and Accuracy.
Precision is very important to customize and personalize the outcomes. Especially in the fields of Medicine, Engineering and Agriculture. Technology advancements when combined with Deep Learning techniques, can make the systems more intelligent and create analytics that are more specific and accurate to a person.
Precision agriculture:
Farm beats project of Microsoft indicates the scope of improvement possible in this fields which can be addressed by the Data Science and Analytics. Using sensor data, this project aims at applying water/pesticides only at the locations where they are needed.
Originally published at https://www.meritedin.com.
