How to get both texts and conditions proposed?
Because our algorithm also thinks to know some interesting conditions
Squadra Machine Learning Company
Last Update 3 yıl önce
In the previous article, it is shown how to train our algorithm in such a way that it is capable of proposing texts by the feature(s) and condition(s) you have selected. However, once enough interesting trainingsdata (example texts) are uploaded, the algorithm is also capable of creating conditions as well.
Within every template, there is a standard section that is called: 'Generate conditions with texts for'. By opening this section, features are shown where the algorithm is capable of creating conditions with related texts.
Keep in mind, this list of features is very likely shorter than the whole list of features where texts can be proposed. Compare this with a human mind again, some features simply can't be grouped or divided by certain values.
Once you select one of the features, it will automatically generate texts that apply to the created condition.