What Kind of Data Do You Want to Use to Train Your Model?
You can train an Image Classifier model by selecting Images from the Datatype Selection panel.
Upload your images belonging to one category all at once, and then update the Label under which you want this to get uploaded. This will help our system to understand its category.
Upload Images for First Category
Consider a sample of Celebrity classification model, where we try to classify different Celebrities. Now, upload images of any one Calebrity under one Category eg. Christiano Ronaldo.
Add Category- After the uploading is done, give a label to the uploaded dataset and hit Add Category
After adding the category, your page will look something like this:
Upload Images for Second Category
For our same Celebrity classification model we need to upload the images for second Category, here we are taking Christian Bale's (Yeah, Batman) images.
Add Category - After the uploading is done, give a label to the set and hit Add Category
You atleast need to add 20 images to each category for the Image Classification Models to work
Now you have added two categories, as shown below:
You atleast need to have 'Two' categories for the Image Classification Models to work
Keep Uploading images for different categories (If you want)
If you want to build an Image Classification model for many different categories then you can add as many categories you want. Just follow the steps mentioned above and click
Next Button once you are done with uploading the images.
You can start building a a Text Classifier model using text option from the Datatype selection panel.
You can train your Text Classifier model using data from one of the three given options below.
URL- Provide a URL containing the desired “*.txt” files. (Coming soon)
Local- Upload “*.txt” files from your device’s hard drive.
Manual- Manually enter your text using this option. (Coming soon)
In the case you’re uploading an Unlabelled .txt file, select the Label under which you’d like this to get uploaded. This will help our system to understand its category.
Adding categories and labels is same as shown above in Image Classification Models (Refer above).
You atleast need to have 'Two' categories for the Text Classification Models to work
Assign Column Labels (Coming soon in v2.0)
While uploading “.csv” files be careful to select the correct column. You’ll have the option of selecting different labels for respective columns. There can be only One Label Column and only One Id/Serial Number. If you think there are more than One Sample Columns, label it accordingly. More on CSV coming soon.
Twitter (Coming soon. Stay tuned for v2.0)
You can train Sentiments, Semantics, Text Classifier and much more using tweets from Twitter too.
Choose between Keyword, Hashtags, Mentions, URLs and filter it to your liking. Separate two or more keywords using “and” or “or” to form a relationship between two different entities.