Preparing Your Data for AI Zwei
1. Preparing Structured Data
For structured data, AI Zwei-friendly formats include files with csv, tsv, xlsx and xls extensions. In addition, Google Sheets can also be used. To prepare your data, make sure that the names of your variables (or features) are given in the top row of your Excel sheet, Google Sheet or text file, one variable name per column. Each example, or instance, should appear in the rows below this, one example per row. Make sure the values for each example appear in the correct columns. This arrangement can be seen in the example below for a structured dataset about diabetes in the Pima Indian community. In this example, the top row in which variable names are given is highlighted in orange.
Once your data are prepared, you can either attach the file to your response to AI Zwei when prompted to upload your data, or you can serve your file on a public URL and simply share the link to the URL.
2. Preparing Image Data For Classification
To prepare image data for a classification problem (e.g., a problem for which you wish to build a model that will automatically identify an image as either that of a cat or that of a dog), simply create a folder for each unique label (e.g., one for "cat" and one for "dog"), naming the folder accordingly. In each folder include all images pertaining to that label. Place these folders in on otherwise empty public url, and simply share the link when prompted by AI Zwei to upload your data!
[Please be patient with us. Instructions for a multititude of data types coming soon]