Wednesday, 7 September 2011

Learnings from Business Analytics

Business Analytics focuses on identifying the changes to an organization that are required to achieve strategic goals.These include changes in strategy , structure, processes , policy and information system.

It focuses on understanding the needs of the business as a whole , its strategic direction and identifying initiative that will help make those strategies into goals.

It includes creating and maintaining business architecture , feasibility studies , identifying new business opportunities, preparing business case and conducting the initial risk assessment.
Here I try to sum up what I have learnt till now in the BA course.

Factor analysis
Factor analysis is a collection of methods used to examine how constructs influence the responses on a number of measured variables.
Measures that are highly correlated are likely influenced by the same factors while measures that are not highly correlated are likely uninfluenced by different factors.
Factor analysis has its application in various fields like economics (to get important factors which affect the economy) , Behavioural Analysis(What are the important factors impacting the human
Behaviour), Marketing(What are the major factors impacting the decision of a consumer to select a specific brand.
It can also be used in market research. Market research data can be factored to understand the results more clearly.

Discriminant Analysis:

It is most often used to help a researcher predict the group in which a subject belongs.
The predictor variable needs to be ordinal or scale, which helps us analyze the effect of such variables on our hypothesis. The most famous application of this method is bankruptcy prediction where Altman z score is used which tells us whether a firm will survive or not.

Conjoint Analysis:

It applies a complex form of analysis of variance to the data obtained from each respondent.
Then it calculates the value for each feature. Features with the highest value are judged the most important to respondents. It tries to identify the interdependency existing between a no. of variables.

It can be used to investigate the attributes that influence individual investors decision making process to buy shares. It also tells us how people make choices between products or services, so that companies can design new products or service meeting customer needs.

Cluster Analysis:

K-means clustering:
It is an algorithm for clustering data that allows the user to choose the number of clusters one would like to have from a given set , based on some similarity.
With a large data set K-means is faster than hierarchical clustering. Data reduction is accomplished by replacing the co-ordinates of each point on a cluster with the cluster’s centroid.
K-means can be helpful for a bank , which is looking at analysing customers in the corporate world on who to target and how much loan to offer to which kind of customers. K-means clustering can be used to find the ideal company with the best risk –return profile using the ratios( like cash , inventory turnover, ROE, ROA) and then cluster of companies around the ideal company can be created. All the companies in that cluster can then be suitable for the bank.

Hierarchical clustering:
Hierarchical clustering is nothing but grouping of data on various scales. Usually a cluster tree is created , which is also known as a dendrogram. In dendrograms, the first step is to find which elements should be merged in a cluster. For the same purpose, two elements which are closest are taken and the same process gets repeated.

Posted by
Neeraj Singh

No comments:

Post a Comment