Build a Machine Learning model in Minutes.

Published in

TEK Society

4 min readOct 6, 2020

Machine learning for everyone…

Whether you’re a coder or not, You can build your own Machine learning model from scratch on your browser in just 10 minutes.

Machine Learning is one of the most demanding skillsets today. It is a subset of Artificial Intelligence (AI) that gives machines the ability to gain knowledge and make improvements from past data and experience.

Machine Learning is a complex technology which has some troubles while using like:

Inability to Analyze Data.
Choosing the correct framework.
Availability of many approaches for the same solution.

Google has simplified the process even further by letting users train their models with the click of a button, no coding required, and export it to websites, apps, physical machines and more.

Teachable Machine(TM) is an AI experiment by Google. It is a web-based tool that makes it fast and easy to create machine learning models without any expertise or coding accessible to everyone.

What can you do with TM?

Users can train a simple model with the use of images, audio and also video clips without any coding required as datasets for training.

And for all of this you just need setup of a computer & webcam.

Steps to use

Gathering Data: Gather and group your examples into classes, or categories, that you want the computer to learn. Upload your own image files, or capture them live with a mic or webcam.
Train the model: Click on- train model, tensorflow.js starts training a neural network in your browser.
Testing model: There you can see your model output and done.

Now If you want you can also export your trained model with a click on Export Model.

Time to do some practical

Let’s make the project(ML model) which detects that the person/or teddy is wearing a mask or not.

Click on Get started and choose Image project :

Gathering the data: Edit class 1 name to mask detected and when you are ready with mask hold the button hold to record to take image data to detect mask. Similarly edit class 2 to Unprotected teddy {choose name by yours} and when you are ready without mask, Hold to record and release after taking the samples.

2. Training Model: Now you are ready to train the model. You can also change the epochs and batch size by scroll down to advanced option to get better accuracy.

Epochs is a neural network term means how many time each and every sample of the dataset is fed through the training model. Higher value of epochs better the model trained.

one epoch = one forward pass and one backward pass of all the training

Batch size the number of training examples in one epoch pass. The higher the batch size, the more memory space you’ll need.

3. Test the model: The model is ready. You can check the accuracy of the model and test it in the preview section. If you are not satisfied with the results you can adjust the settings while training model.