Recherche : [MachineLearning] - Zestes d'idées // Ideas zests

317 links
28 private links

Links per page: 20 50 100

6 results tagged MachineLearning x

hyperparameter_hunter

Wrapper pour la calibration de modèles de ML et consolidation de résultats
//
ML models wrapper and results consolidation
August 28, 2018 at 10:22:21 PM GMT+2 - permalink -

- - https://github.com/HunterMcGushion/hyperparameter_hunter

MachineLearning Wrapper Tool
Feature importances with forests of trees
May 1, 2018 at 1:26:58 AM GMT+2 - permalink -

- - http://scikit-learn.org/stable/auto_examples/ensemble/plot_forest_importances.html

Scikit-learn Python MachineLearning
Comparison of Calibration of Classifiers
May 1, 2018 at 1:25:55 AM GMT+2 - permalink -

- - http://scikit-learn.org/stable/auto_examples/calibration/plot_compare_calibration.html#sphx-glr-auto-examples-calibration-plot-compare-calibration-py

Scikit-learn Python MachineLearning
The Building Blocks of Interpretability

Recherche sur l'interprétabilité des réseaux de neurone profonds
//
Deep learning interpretability research
March 9, 2018 at 4:02:37 PM GMT+1 - permalink -

- - https://distill.pub/2018/building-blocks/

MachineLearning DeepLearning Visualisation
Luminoth : an open source toolkit for computer vision

Détection d'objets et classification d'images, le tout en Python en utilisant TensorFlow et Sonnet
//
Object detection and image classification in Python using TensorFlow and Sonnet
//
https://www.tensorflow.org/
https://github.com/deepmind/sonnet
October 12, 2017 at 8:52:55 PM GMT+2 - permalink -

- - https://github.com/tryolabs/luminoth

Python MachineLearning PictureAnalysis DeepLearning NeuralNetwork
La reconnaissance d'image expliquée par les pokémons

Une introduction à la reconnaissance d'images via apprentissage supervisé
//
Introduction to image recognition through supervised learning
March 4, 2016 at 10:50:46 AM GMT+1 - permalink -

- - https://medium.com/@dimart/pok%C3%A9mon-recognition-d3ad5cadc61e#.jjzblpx7w

UsefullAndFun DataAnalyse MachineLearning

Links per page: 20 50 100