Deep Learning Lab Create Tab

The Create tab is the place to create a new Machine Learning Model.

The Create tab contains a link (1) to a help screen with information about how to use this tab.

Create Input Worksheet The Create button (2) will add a new worksheet to Excel.

Use this worksheet to input the training data for your new model.

The first row of the worksheet should contain headers for your data. The last column MUST contain the target variables you want your model to predict.

Analyze Data for Model

When you select the Analyze Data button, the Deep Learning Laboratory (DL Lab) will review the training data, determine the type of data in each column, and calculate various statistics for the data. The DL Lab adds a row of Data Type Selectors just under the column headers with its best guess (Continuous or Categorical) for the type of data in that column.

The Continuous data type can take a range of values such as monitory value or temperature. The values may be positive or negative. Categorical data indicates the data belongs to different groups or classes. Examples are a person's gender, or the grades given on a test. It's important that you review and make any changes to these data types.

The Deep Learning Laboratory will create a new worksheet named DataSummary, which contains a statistical summary of the continuous data in the TrainingData worksheet.

The Deep Learning Laboratory will also create a new worksheet named ModelSummary, which contains a summary of the number and size of the layers in the deep neural network.

Create and Save Model

DL Lab will let you choose the number of layers and the size of each layer (3). The default model size is 2 layers of 50 neurons each. Be sure and use commas to separate your layers.

Give your new model a name and a short description and then save your model. After saving the model it will become the active model in the Deep Learning Laboratory.