Week 5 – Data Mining and Neural Networks

Dr. Ilona spoke about the various techniques about data mining that including decision trees, neural networks it was really interesting to know how companies use them in real time to improve or maintain their customers database. Many approaches are used but in the end it seems they get to almost same result.
It was quite interesting to know that they are used in financial industry where credit provided based on its decision and humans. I still wonder the accuracy of the neural networks and if they are down to earth rather than following the written algorithm. Although these system are thoroughly tested, at a one point of time depending much on neural networks seems too “technological “rather that really understanding and assessing customer needs. The example given by Ilona where Mobile Phone Company uses to segment and identify what customer needs are and try changing approach, in this example the company uses it for back end monitoring and analyzing results and not to decide the customer’s ability. Artificial intelligence as the name states it is Artificial and not natural. The companies as of now at least relay too much on computers. This also leads to fraud creeping in. and again I believe AI or neural networks can’t do anything to prevent them unless a human identifies and corrects it. I do agree with the video that discussed some ethical issues like invading privacy. If companies get to access all the information (though customer’s consent is obtained but I believe that is just too good) in the past and predict the future, then that is inacceptable. A simple example of this if a customer applies for credit of company X and declined, he applies at company Y and approved, what justification does this give?. I assume either one fo them should be wrong or both the companies might be wrong.
To conclude using AI and neural networks should be balanced and used in such instances where it will help to get the most positive influence in customer rather that just adopting for all the processes. This in a long run will help companies to understand the customer’s needs and serve them better which will lead to strong relationship with customers.
Venkat 20105169