Back

Article

What is Small Data and what to do with it?

1

How to work with small datasets and what is the difference between Big Data and Small Data.

What is Small Data?

Small Data is data that is most useful, easily accessible, and beneficial to a particular department of an organization, it is rarely a centralized data owned solely by the IT department.

Small Data and Big Data work together and do not compete with each other.

But to better understand the differences, a look at the following table may help:

Know What You Have Access to – Do a Small Data Audit!

Today data is considered the lifeblood of an organization, though you can’t make it work for you until you know what data you possess, the quality of your data and its quantity.

The good news is – Small Data is everywhere. Every department has small data but how to know exactly where to find it and how to extract the maximum value of it?

To answer this question, find out what you have by making a small data audit.

To start, make a list of all the different types of data you might have. The list will differ from one industry to another but below are the examples of possible data:

  • CRM software customer information
  • Purchase information about raw materials, equipment, marketing materials etc;
  • Online shopping cart data
  • Sales by customer and by product/service
  • Server with customer info in Excel
  • Behavioral data from the website
  • Data from a machine
  • IoT data
  • Performance data, etc

Once you’ve listed the types of data, you can now follow these steps for your small data audit:

  • Find out where it is
  • Interview the key players
  • Prioritize and organize
  • Track how your data is being used

How to work with Small Data?

There are several problems of working with small datasets that mainly revolve around high variance:

  • Overfitting: occurs when a model tries to predict a trend in data that is too noisy. A model that has learned the noise instead of the signal is considered “overfit” because it fits the training dataset but has poor fit with new datasets.
  • Outliers: small amount of data that differs a lot from the majority of data and deviates the average number as a result;

So, what to do in this case?

The easiest and best way to work with small datasets is to use the technology of MyDataModels that works exceptionally well with Small Data.

Here are a few other advantages of MyDataModels technology:

  • It is easy to use and does not require training, coding skills nor the knowledge of machine learning. It has been created with the target to democratize machine learning;
  • It’s fast and user-friendly. It takes less time to build models with MyDataModels technology comparing to other software (hours versus months) and the platform is easy to use;
  • It can be used on a laptop but as well on a desktop, cloud and mobile;
  • It offers a free account and other packages according to the specific needs of each user.

You can get started with MyDataModels now, it’s free.

Create your free account or learn more about MyDataModels packages