Quantitative Analysis

Quantitative analysis has been used in ancient India and around the world for thousands of years. The decision makers in business and industries have been using this techniques to gain profit and for constant growth.

But what is Quantitative Analysis exactly ?

Quantitative analysis is a technique that uses basic Mathematical and Statistical modeling, measurements and research to understand the behavior of the the data variables. A procedure where a raw data is processed and manipulated to produce meaningful information. Quantitative analysts represent a given reality in terms of a numerical value. So when we talk about quantitative analysis, one should also know about qualitative analysis. What is the difference between them.

• Quantitative techniques are those statistical and programing techniques which helps decision makers solver many problems, especially those concerning business and industry. Factors: {Different Investment Alternative, Interest rate, Inventory level, Labor Cost, Demand}
• Qualitative techniques are those techniques that provide the decision makers with systematic and powerful means of analysis, based on quantitative data, for achieving predetermined data. Factors: {Breakthrough in technology, Weather report, etc}

Quantitative Analysis Approach:

1. Defining the problem : This is most import and difficult step, because what you may analyze can also be just a symptom and not the real problem. It is very much necessary to go beyond the symptoms and get a clear understanding of the true cause. A clear and concise statement can give direction to subsequent steps to solve the problem. Some times it may be a necessity to concentrate on few of the problems, hence selecting the right problem is very import. Lastly specific and measurable objective have to be developed.
2. Developing Model : Math models are very common and popular, however quantitative analysis model are very realistic, solvable and understandable mathematical representation of the situation. There are many different types of model like "Scale Model", "Schematic Model", Stochastic Model", etc. In these models there are always two types of variables, controllable and uncontrollable also called independent and dependent variables. The dependent variable changes it behavior when the independent variables show some changes in itself.
3. Acquiring Input Data : Acquiring data is most important step. Even if we have the best of our model and if the data is not accurate, it is obvious that the result will be un-satisfying. Data may come from various sources such as company report, operations department, company document, interviews, surveys, on-site direct measurement or statistical sampling. However, the thumb rule one should follow is "GIGO" - 'Garbage In Garbage Out', which simply means, if you feed garbage data to the model it will give garbage result.
4. Developing a solution : There are many methods to develop solution after once you have got the data and model, example you can find the optimal values for the model, so you can optimize a model and find optimal parameters that will yield desired outcome. There are 'Trial & Error' methods, if the model is very complex for example, you may keep trying different design till the time you don't get the desired outcome. And another approach to finding a solution is "Complete enumeration". So complete enumeration is basically hundred percent inspection and with the help of computer one can go through all the possible solutions. Solutions can also be derived by using an algorithm, which is a series of repeating steps to reach solution.
5. Testing solution : I believe, that both input and the model should be tested for accuracy before analysis and implementation. New data can be collected to test the model to check the accuracy of the desired output. The result derived from the model should be logical, consistent, and represent the real situation.
6. Analyzing result : When determining the implication of the solution, implementing results often requires changes in the organization. Hence the impact of the action or changes needs to be studied and understood before implementation. Sensitivity analysis determines how much the result will change if the model or input data changes. Suggestion, the sensitivity model should be very thoroughly tested.
7. Implementing result : Implementation incorporates the solution into the company. However, implementation can be difficult, people may be resistant to changes. Many quantitative efforts usually fail because many a time's a good workable solution is not properly implemented.

That's all about it.!!

A Data analyst can easily implement all the steps above to do the analysis using any application of expertise.

Will further update this article with real time use case with python

Till then adios.