{"id":1223,"date":"2023-05-22T13:33:13","date_gmt":"2023-05-22T13:33:13","guid":{"rendered":"https:\/\/catedramasmovil.uc3m.es\/2023\/05\/22\/from-neural-networks-to-decision-trees-uncovering\/"},"modified":"2024-06-27T17:41:25","modified_gmt":"2024-06-27T17:41:25","slug":"from-neural-networks-to-decision-trees-uncovering","status":"publish","type":"post","link":"https:\/\/catedramasmovil.uc3m.es\/en\/2023\/05\/22\/from-neural-networks-to-decision-trees-uncovering\/","title":{"rendered":"From neural networks to decision trees: uncovering\u2026"},"content":{"rendered":"

[et_pb_section fb_built=”1″ _builder_version=”4.17.0″ custom_padding=”0px||||false|false” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_row _builder_version=”4.17.0″ _module_preset=”default” custom_padding=”0px||||false|false” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_column type=”4_4″ _builder_version=”4.17.0″ _module_preset=”default” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_gallery gallery_ids=”1214,1231″ fullwidth=”on” _builder_version=”4.25.1″ _module_preset=”default” hover_enabled=”0″ global_colors_info=”{}” theme_builder_area=”post_content” sticky_enabled=”0″][\/et_pb_gallery][\/et_pb_column][\/et_pb_row][et_pb_row _builder_version=”4.16″ background_size=”initial” background_position=”top_left” background_repeat=”repeat” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_column type=”4_4″ _builder_version=”4.16″ custom_padding=”|||” global_colors_info=”{}” custom_padding__hover=”|||” theme_builder_area=”post_content”][et_pb_text _builder_version=”4.18.0″ _module_preset=”default” global_colors_info=”{}” theme_builder_area=”post_content”]<\/p>\n

<\/h2>\n

Objective<\/h2>\n

<\/span><\/p>\n

In this research project we have developed an algorithm capable of transforming pre-trained neural network models into tree structures.<\/div>\n
<\/div>\n
Despite the great generalization power of neural networks and their success in solving complex problems in many industries, they render inadequate when the problem requires, in addition to a solution, an explanation of the decisions taken to reach such conclusion. This fact makes them unsuitable for use in sectors that require interpretability, either due to legal or commercial restrictions.<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
For this reason, during the project we have worked on an alternative interpretation of these models that allows us to extract features from the attributes without any loss of generalization of the original model during the process, contrary to other similar attempts. The tree representation allows, on the one hand, to represent the network output as a decision making sequence based on the values of the input variables and, on the other hand, to carry out an analysis of the local linear models that are generated by the network in order to apply statistical techniques to them.<\/div>\n
This lays the groundwork for a deeper analysis of the network structure and the application of different techniques to exploit this alternative representation.<\/div>\n

[\/et_pb_text][\/et_pb_column][\/et_pb_row][\/et_pb_section][et_pb_section fb_built=”1″ _builder_version=”4.18.0″ _module_preset=”default” custom_margin=”20px||||false|false” custom_padding=”0px||||false|false” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_row column_structure=”1_2,1_2″ _builder_version=”4.18.0″ _module_preset=”default” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_column type=”1_2″ _builder_version=”4.18.0″ _module_preset=”default” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_image src=”https:\/\/storage.googleapis.com\/wp-uploads.bucket.wp.uc3m.es\/wp-content\/uploads\/sites\/70\/2023\/05\/22131739\/6D6C371A-14FB-48C3-A3B8-B8C29950DCF2_1_105_c.jpeg” title_text=”6D6C371A-14FB-48C3-A3B8-B8C29950DCF2_1_105_c” _builder_version=”4.25.1″ _module_preset=”default” hover_enabled=”0″ global_colors_info=”{}” theme_builder_area=”post_content” sticky_enabled=”0″][\/et_pb_image][\/et_pb_column][et_pb_column type=”1_2″ _builder_version=”4.18.0″ _module_preset=”default” global_colors_info=”{}” theme_builder_area=”post_content”][et_pb_text _builder_version=”4.25.1″ _module_preset=”default” custom_padding=”||0px|||” hover_enabled=”0″ global_colors_info=”{}” theme_builder_area=”post_content” sticky_enabled=”0″]<\/p>\n

BACHELOR’S THESIS BY:<\/strong><\/span><\/p>\n

ERNESTO VIEIRA MANZANERA<\/strong><\/span><\/p>\n

[\/et_pb_text][et_pb_text _builder_version=”4.25.1″ _module_preset=”default” global_colors_info=”{}” theme_builder_area=”post_content”]<\/p>\n

Academic Experience<\/strong><\/strong><\/p>\n