Continuing to my previous blog, let’s look into some of the data science applications to understand data science better.
Let us see how the world is benefiting a lot from the data science, whether it is the field of retail, banking, healthcare or personal uses. The evolution of computer processing capabilities and open source data analysis software and combining with the predictive modeling, is has become very affordable and easy for commercial purpose.
Data science applications in finance.
Fraud detection: Credit Card fraud detection is a very common area where data science is used. What usually happens that the credit card fraud exhibits a pattern and relationship between different entities.
With the help of historical data and the machine learning algorithms this pattern is identified and amodel is build to be used in the live transactions.
Now when a new transaction takes place the model analyse all the attributes and provide the fraud score which is nothing but the probability of the transaction to be fraud. With the threshold defined in the model the transaction is classified into fraud or actual.
If identified as fraud transaction the bank immediatly takes action, some times the transaction is stopped, sometimes the creditcard is blocked untill the verification is done for the credit card owner.
With machine learning in place the prediction has become more accurate and it not only saves fraud transactions but also reduces the cases of holding correct transactions which cause a lot of harassment to the customer if the transaction is geniune.
Statistically speacking, with the help of machine learning not only 15% of fraud detection is increased but almost 50% of false alarm has also reduced resulting 60% of savings.
Data analytics in retail industry.
Retail predictive analytics is used for forecasting, inventory management and assortment planning in retail domain using the historical data for past 2-3 years.
With the traditional time series analysis, forecasting could be done but that is not helpful if multiple attributes are added then combining time series with machine learning gives the best result.
Some other retail analytics use cases are for basket recommendation. You see when you open any online site like amazon you see they immediately start providing you recommendations.
They usually study the buying pattern of a lot of users and your ethnicity, age, gender and previous buying patterns based on that with the help of association predictive analysis they recommend you.
Health data science.
Healthcare data analytics is another major field where data science ha bought fantastic results. How healthcare data scientist uses data science.
Data analytics in healthcare industry can be used for identifying the diseases more accurately based on the different results/symptoms of previous patients and more effective medicines can be provided for the cure.
Another way is in preventive analysis, predicting diseases outbreaks and alerting the concerned authority to take preventive measures to prevent it or readiness to cure it.
This can be done by looking at what people are searching in google or what people are twitting a twitter. Here data is collected from the google search, twitters API and sources like that and since these data always has the location information associated with it.
Healthcare data scientist keeps on monitoring these and when they find any pattern that a certain disease is searched from the location a lot then there might be a chance of a diseases outbreaks.
These are some of the interesting data science applications which must have provided you some clarity about the data science in depth.
In the subsequent blogs I will discuss more about the data science life cycle and other skill set required to become a data scientist.
Till the time please do not forget to provide your feedback.