Add simple_classify Method to DecisionLayer Class #1
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Added simple_categorise which uses sum of Cosine Similarity Scores to determine Category. Option to use tan function to boost scores for closest points, and reduce scores for further away points.
It now takes a query result as an argument and outputs scores_by_category too, for debugging purposes.
…bs/semantic-layer into simple_classification
…bs/semantic-layer into simple_classification
Test cases aren't just the original Decision utterances now, but semantically similar utterances. Also added new max_score_in_top_class method, which chooses the top score of the top scoring vector in the top class to compare to the threshold value.
Threshold checks done outside of _semantic_classify. Testing more efficient as not using dl._query() accross every threshold.
These now match the number of non-other types.
jamescalam
approved these changes
Nov 9, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description:
This pull request introduces a new method, simple_classify, to the DecisionLayer class. The simple_classify method takes a dictionary of query results and an optional boolean parameter apply_tan (defaulting to True), and returns the category with the highest total score, along with the scores of all categories.
The method works as follows:
It applies a scoring system to the results from the _query method. If apply_tan is True, the score for each result is calculated as the tangent of the result's score times π/2. If apply_tan is False, the raw score is used.
The scores are grouped by category. If a category appears multiple times in the results, the scores for that category are added together.
The categories are sorted by total score in descending order.
The method returns the category with the highest total score. If there are no results from _query, the method returns None.
This new method allows us to categorize a query based on the scores from _query, with the option to apply a tangent transformation to the scores. This can be useful when we want the scores to increase exponentially as the cosine similarity increases.