5 Easy Steps To Get Started In Machine Learning On Teradata Aster

The best minds from Teradata, our partners, and customers blog about relevant topics and features.
Teradata Employee

How do I get started with Machine Learning? 

Where should I begin, if I do not have a background in math?

Photo Credit

The Long road 

I am thinking to learn a tool for a month, then learn machine learning theory for another two months and start implementing the projects

This is precisely the bottom up approach towards machine learning. This approach would certainly work, but you would not reach your expected results soon. The entire topic of machine learning could be overwhelming with umpteen number of blogs and courses churning content after content. Most content use a bottom up approach of learning the math behind , understanding different algorithms, learning new languages and at last learn machine learning.

This blog post would address a step by step top down and fail fast approach to applied machine learning. Gear up to get started, but remember..

Bug Taking Little steps

When in Doubt, Choose Select *

SQL is the simple and easiest form to express analytic. Having to go through a new language could involve a steep learning curve. Also most developers know SQL to start with and it provides quick and easy access to the data.

Get and Load Data

UCI Machine Learning Repository  is a good repository of real world examples. This would be a best place to start with. Choose and load the data set into Aster. Start with the simple Iris Data Set.

Get your Hands Dirty

Understand the process behind solutions to ML problems. In the post Decision Trees with Teradata Aster

have laid down the most common steps of the machine learning process.

  1. Define the problem: With the dataset in hand, focus on clearly defining the machine learning problem to be solved.
  2. Analyze data: Analyze the input features for patterns and information that could be used to develop a model.
  3. Data preparation:  Preparing the analytic data set gets you closer to applying your model. Aster comes with a good set of transformation functions to start with. Please follow the steps highlighted in the post above to go through the steps in detail . Explore few of these functions and prepare the data for modelling.
  4. Model: Just choose the simplest algorithm to start with. You do not need to understand how it works to use it in the first place. In the above post I have highlighted the use of decision trees
  5. Evaluate:  Evaluate the performance of the model and choose the best performing model.

Practice, Practice, Practice

Congratulate yourself for having successfully completed the first machine learning workflow on Aster. But the key to Mastery is deliberate practice. it is important to rinse and repeat this process on different dataset across different domains.

Shifting Gears

Now that you have implemented your machine learning process using Teradata Aster. The next step is to shift gears and understanding what is happening. In the post Decision Trees with Teradata Aster have briefly explained how decision trees work. Diving deep to understand the algorithm and implementation would help you gain more knowledge.


In this post we saw the 5 easy steps to get started with machine learning on Teradata Aster.

If you liked this post and would like to learn data science - do follow my posts on learndatascience.net

for more information on Data Science.