- Data feasibility study

Description

It is the purpose of a data feasibility study to determine whether your available data can be used for machine learning and artificial intelligence.

By performing a feasibility study, you can clarify what potential exists in your data and how you can use it to gain greater insight into your company and help you make better decisions.

Additionally, it will clarify any possible errors that may be found in the data, as well as what can be done to correct them. Which can be divided into two categories: collecting additional data or improving the quality of data that has already been collected.

Requirements for data feasibility study

The first requirement for a successful data analysis is that the company has already gathered data. It is the goal of the feasibility study to determine the quality of this data. Having access to data is essential for determining the quality of our testing, as we can test what happens during testing.

Similarly, it is crucial that the method of collecting the data is thoroughly described so that we can identify areas where the quality of the collected data can be improved or discover new methods of collecting data.

Technical possibilities with a data feasibility study

The technical potential of a preliminary research study is that you are subsequently able to determine what type of AI and machine learning your data is capable of supporting.

In essence, these concepts cover a wide range of techniques that are all intended to uncover trends in the data that you have, where the feasibility study clarifies which options on the list have the greatest potential given the data at hand. These include classification, regression, suggestion algorithms, Natural Language Processing, and image recognition.

What suits you best depends on the data you have available and what you would like to have anticipated.

The process regarding a data feasibility study

Q-analytics process when we do a feasibility study is:

  • Initial meeting
  • Workshop
  • Data access
  • The analysis phase
  • Result

Initial meeting

During the introductory meeting, we will discuss scoping for a preliminary investigation, what some goals you have in relation to AI/ML projects you plan on realizing in the future, and a general overview of what type of data you have available.

Workshop

During the workshop, you will get a precise overview of the data you have, as well as the description of the data. Our aim is to get all the information ready for the pre-analysis and convey it at the workshop. 2-3 hours.

Data access

In this phase, we gain access to the data. We could receive the data and then conduct a preliminary investigation in our office. A preliminary study can also be done locally with you if the data is especially sensitive. The time period here depends on the data we need to access and how we need to access it. 2-4 hours.

The study itself

Here, we will conduct preliminary research and create the finished product for you. It usually takes between one and two weeks, depending on the scope and the objectives.

Result

The final product will be presented to you, along with an explanation of what we have found. This takes approx. 1-2 hours.