AI experiments advanced options

<< Click to Display Table of Contents >>

Navigation:  » No topics above this level«

AI experiments advanced options

Overview

Once you have created an AI experiment, generated the model, and are evaluating the accuracy of the prediction, you may switch to advanced options in order to modify the algorithm employed or explore other combinations.

These options generally assume a good degree of expertise in machine learning.

For information about generating the model to interpret results, refer to Working with AI experiments.

 

Advanced options

Switch to the advanced options by clicking Advanced View:

 

Cloud_AI_switchdvanced

 

In the advanced view, you see a graphical view of how Bizagi Artificial Intelligence has determined the best treatment and algorithm for the predictive service.

The view presents the steps carried (each step represented by a box) in process order from top to bottom.

 

There are two different flows you may want to reconfigure:

 

1. The training flow.

This handles data treatment, algorithm choice and training of the model, as employed by Bizagi during the model generation step.

Toggle the training flow on or off by clicking the icon having with two arrows:

 

Cloud_AI_expert_ds

 

2. The predictive flow.

This is the series of steps Bizagi uses at runtime, whenever predictive analysis service is invoked.

Toggle the predictive flow on or off by clicking the icon with two arrows:

 

Cloud_AI_predictiveFlow

 

Interpreting the flows

You can explore the sequence of each flow and review the specific configuration for each of the steps by clicking its box.

To the right, you will see the configuration Bizagi Artificial Intelligence has applied as being best suited for the use case.

 

The image below shows an example of a training flow and how the first step is configured.

The step is of the type DataSource (typically the first one for a training flow), the step where data is obtained.

It is configured with indications of the target Dataset and the attributes to use (selected features):

 

Cloud_AI_interpretds

 

By clicking Feature selection for this type of step, you may redefine the selected features, as described in Working with AI experiments.

 

For other steps, such as Clean Data, different settings are available.

The image below shows as an example, that the upper Clean Data step applies for the pregnant, plasma, bpressure, skinthickness, 2h_insulin, BMI, and age attributes and that its clean policy for empty values replaces them by the average value of the set.

 

Cloud_AI_interpretclean

 

You may see more than one Clean Data step. Each step would´provide a different treatment for empty values for the selected attributes.

 

For replacing empty values, you have options according to the data type of the feature, as shown in Working with AI experiments.

 

Cloud_AI_expert_trainingflow2

 

For other steps, there may be multiple sequence lines running from them, or to them.

In the case of Split Data, a left fraction indicates what fraction of data goes to one step (and by implication, the remaining fraction that goes to another step).

The image below shows how Split Data sends 70% of its data to the Train Model step (to achieve a best certainty for a prediction, a preselected model is used), and the remaining 30% goes to the Validate Model (to verify test results).

 

Cloud_AI_interpretsplit

 

You have also a step indicating which algorithm is selected for the model. The image above shows Multiple Linear Regression.

In a similar way and usually fewer steps, a predictive flow is interpreted.

 

Modifying the flows

To modify how a flow is defined, use the boxes in the left panel, as grouped per their function.

To include any of these in the flow, simply drag and drop them into place:

 

Cloud_AI_expert_trainingflow

 

To delete a step so you can connect other ones, select a step and use the keyboard's delete key.

Then use the new shape’s endpoints by dragging a sequence line from the source shape to an endpoint on the next shape in the sequence:

 

Cloud_AI_expert_trainingflow5

 

Connect to the target shape's endpoint:

 

Cloud_AI_expert_trainingflow4

 

In the end, make sure you have redefined the flow so it is consistent and adequately connects all steps in the proper order:

 

Cloud_AI_expert_trainingflow3

 

You can use Model evaluations such as: Score model, Validate model, or Cross validation:

 

Cloud_AI_expert_modelev

 

You can use Data Transformations such as: Word count, Random split, SMOTE, Select Columns, Remove Columns, Normalize Numbers, Make categorical, Join Data sets, or Clean missing values (Clean Data):

 

Cloud_AI_expert_datatransf

 

You can use Machine learning algorithms such as: C45, ID3, Linear SVM Classifier, Multiple Linear Regression, Logistic Regression or Train Model:

 

Cloud_AI_expert_mlalgorithms

 

Finally, when you are done, generate the model and make sure you review its results and test it thoroughly. Then, if the experiment is satisfactory, you can publish it, as described in Publishing an AI experiment.