by   

 

 

Support

How to Use



 

Do you want a more detailed guide?

 

 

 

FAQ


General

 

 

What is the LinkSquare AI Platform?

The LinkSquare spectrometer is a smart handheld spectrometer that will be a unique addition to the spectroscopy industry for its significant innovation in performance, cost, and size. Not only the device itself, we have a powerful machine learning algorithm that make sense of spectral data. You can experience the basics of our machine learning algorithm. The AI Platform provides a DIY applet creation tool by analyzing the samples of each item, distinguishing spectral features, defining spectral identity, creating a database, and making an applet that can be downloaded and viewed as an applet in the LinkSquare app.

 

What do I need to use the LinkSquare AI Platform? Is it necessary to purchase a LinkSquare device?

You can sign up the LinkSquare AI Platform with a valid email address. But the LinkSquare Basic or Professional Package is necessary for the best experience of the LinkSquare AI Platform. Only with the LinkSquare Basic or Professional Package, you can collect and upload spectral data for the machine learning algorithm to use.

 

Can I try the LinkSquare AI Platform?

When you first sign up the LinkSquare AI Platform, we offer a week of free-trial so that you can enjoy Standard Plan free of charge. You can check this at the 'billing' page.

 

I can't sign in after creating an account.

Once you submit all information for sign-up, we send you an activation email to the email address you provided. You can sign in after the email verification.

Plan & Billing

 

 

Can I upgrade or downgrade the plan?

Yes, you can upgrade or downgrade your plan. The prorated current subscription fee will be refunded, and a new plan subscription for a month will be started.

 

Can I cancel the subscription and get refund?

You can cancel the current subscription anytime, but the refund will not be made. When you request cancel, the current subscription will remain until the end of subscription period.

 

What happens if the credit card transaction has failed?

We are going to retry the transaction again, and put your account on hold. This event will be notified to you via email. All the data you have uploaded will be intact, and any on-going machine learning process will keep continue until the end. At the next time you sign in the LinkSquare AI Platform, we're going to ask you to update the credit card information.

Applet

 

 

What are applets?

Applet is a small item you can download and use in your daily life. With the LinkSquare AI Platform, you can create your own applets and download them on your smartphone.

 

I have collected data with the LinkSquare App. How can I use it?

You can collect data with the LS Collector application included in the LinkSquare Basic or Professional package. Data collection also can be done via the LinkSquare App, it is only for allowing the data collection to be done easily and remotely. The data on the app needs to be transferred to the LS Collector application and exported in a proper format.

 

Which file can I upload for machine learning?

The LS Collector application included in the LinkSquare Basic or Professional package has a menu to 'export' spectral data. When you choose 'Stratio Data Type' on the dialog, the final outcome is a zip-compressed file. It is made to be uploaded to the LinkSquare AI Platform for machine learning.

 

I'm a Standard plan user. Can I customize applets?

Unfortunately, no. Applet customization is one of the functionalities for Premium Plan users. The image resources for an applet will be automatically set among the prepared default images. The name of each sample will be the same with the one you set during the data collection.

 

I'm a Premium plan user, and I would like to share my applet to others.

Sharing your applet to your friends or colleagues is a great idea! Before you share your applet to them, ask them to sign up the LinkSquare AI Platform, as you did, and let you know the email address used for sign-up. You can search and add their email accounts at the 'Applet Distribution' drop-down menu. When they sign-in at the LinkSquare App, under 'Settings' -> 'Account' menu, your applet will be shown on their downloadable applet lists.

 

What is the meaning of 'iteration'? How can I know which number I need to choose?

The current machine learning algorithm that the LinkSquare AI Platform provides is based on the repetition. As the number of repetition, which is called 'iteration' here, becomes large, the machine learning result usually gets better. However, the law of diminishing returns can be found here; you may not expect the performance increase after enough number of iteration has been done.

 

Do you provide machine learning algorithms other than classification?

The LinkSquare AI Platform only offers a CNN-based classification algorithm modified to work the best with the spectral data collected by the LinkSquare device. As we are aware of the needs of other types of machine learning algorithms, other algorithms will be offered soon.

 

Why does it take time for my applet to begin training?

The machine learning algorithm run on the LinkSquare AI Platform uses Graphics Processing Unit (GPU) because of the intensive parallel computation. Due to the limited number of resources, your applet may need to wait for others' applets have completed their computation.

 

How can I test the machine learning result?

In order to test the machine learning result, you need to use the LinkSquare App. After you sign in at the app under 'Settings' -> 'Account' menu, your applets will be shown on the downloadable applet list.

 

I'm not a machine learning expert. What can I do?

The LinkSquare AI Platform is for the users to create their own applets. If you think that you don't have enough knowledge about spectral data and/or machine learning, you are always welcome to contact us and discuss the possible collaboration with us.


Any questions about LinkSquare? Send us an email!

linksquare@stratiotechnology.com

×

Some text..

×

Some text..

×

Some text..

×

Modal Header

Some text..